John E. Elicker - Senior Vice President of Public Affairs and Investor Relations Lamberto Andreotti - Chief Executive Officer, Director, Member of Executive Committee and Member of Securities Issuance Committee Charles A. Bancroft - Chief Financial Officer and Executive Vice President Francis M. Cuss - Chief Scientific Officer, Executive Vice President and Member of Science & Technology Committee Giovanni Caforio - Chief Operating Officer and Director

Seamus Fernandez - Leerink Swann LLC, Research Division Gregory B. Gilbert - Deutsche Bank AG, Research Division Jami Rubin - Goldman Sachs Group Inc., Research Division Timothy Anderson - Sanford C. Bernstein & Co., LLC., Research Division Christopher T. Schott - JP Morgan Chase & Co, Research Division Andrew S. Baum - Citigroup Inc, Research Division Mark J. Schoenebaum - ISI Group Inc., Research Division Ronak H. Shah - Crédit Suisse AG, Research Division David Risinger - Morgan Stanley, Research Division Alex Arfaei - BMO Capital Markets U.S. Steve Scala - Cowen and Company, LLC, Research Division Colin Bristow - BofA Merrill Lynch, Research Division Charles Anthony Butler - Guggenheim Securities, LLC, Research Division

Good day, and welcome to today's Third Quarter 2014 Earnings Conference Call. This call is being recorded. At this time, I would like to turn the call over to John Elicker. Please go ahead, sir.

Thanks, Travis. Good morning, everybody, and thanks for joining us to review our Q3 results. With me this morning are Lamberto Andreotti, our Chief Executive Officer; and Charlie Bancroft, our Chief Financial Officer. They'll both have prepared remarks. And then also joining are Giovanni Caforio, our Chief Operating Officer; and Francis Cuss, our Chief Scientific Officer, who will be here for Q&A. So before I turn it over to Lamberto, I'll handle the Safe Harbor language. During the call, we will make statements about the company's future plans and prospects that constitute forward-looking statements. Actual results may differ materially from those indicated by these forward-looking statements as a result of various important factors, including those discussed in the company's SEC filings. These forward-looking statements represent our estimates as of today and should not be relied upon as representing our estimates as of any subsequent date. We specifically disclaim any obligation to update forward-looking statements even if our estimates change. We'll also discuss non-GAAP financial measures adjusted to exclude certain specified items. Reconciliations of these non-GAAP measures to the most comparable GAAP measures are available on our website. Lamberto?

Thank you, John. Good morning, everyone. Overall, we had a very good third quarter. We delivered strong sales across our key brands. Revenues were $3.9 billion, a 7% increase over last year, excluding the Diabetes business that we still had in 2013. Equally important, throughout the quarter, we made significant progress in delivering our pipeline, with important clinical and regulatory milestones and several business development agreements. With respect to Eliquis, the combined efforts of our sales, marketing and medical team contributed to its strong performance around the world this year. And approval for the VTE treatment in both the U.S. and Europe this quarter provide another strong platform for success. The strength of our data and our progress to date supports our continued high expectations for Eliquis. With respect to hepatitis C, we have a clear opportunity in Japan where hepatitis C patients have had limited treatment options. We received approval for our dual, Daklinza+Sunvepra, Regimen in July and launched in September. This Dual Regimen, which was developed to meet the distinct need of a significant proportion of the Japanese patient population, has the potential to play a major role in curing HCV patients in that country. We are only at the beginning, but the initial performance indicators give us reason to be optimistic. We are also off to a good start in Europe, where the European Commission approved Daklinza for use in combination with other products across multiple genotypes and where we have launched in Germany and a few other countries. In the U.S., we have withdrawn our new drug application for asunaprevir while we continue to pursue FDA approval of Daklinza, focusing on distinct patient populations with unmet medical need. We look forward to presenting important data at the upcoming AASLD meeting in Boston. And with respect to…

Thank you, Lamberto. Good morning, everyone. As Lamberto noted, we had a strong quarter. Our key brands did particularly well, with Eliquis, Yervoy, ORENCIA and SPRYCEL each posting double-digit sales growth versus the same period last year. Let me provide a few highlights. Eliquis sales were $216 million. We saw continued strong growth during the quarter and have recently launched our new indication for VTE treatment in the U.S. and Europe. Our expectation is that the new indication will be an important catalyst for continued growth. While our focus will remain on cardiologists, the progress we've made in new-to-brand prescription share for AFib and the VTE indication should help as we expand more into primary care. In the U.S., net sales for Eliquis were $113 million, up 20% sequentially from the second quarter. And AFib, our new-to-brand share among cardiologists, is now approaching 45%. Overall, Eliquis' total prescription volume and new-to-brand volume continues to grow faster than any other NOAC. Outside the U.S., net sales were $103 million, up 34% from the second quarter. Yervoy sales grew 47% to $350 million. Yervoy had its best quarter since launch and the recent weekly sales trends remain strong. U.S. sales were $191 million, up 47% from the same quarter last year and up 10% from the previous quarter. We continue to make significant gains in first-line use among immuno-oncologists as they become more comfortable treating patients with Yervoy. Sales outside the U.S. totaled $159 million, also up 47% over last year's quarter, and we continue to see strong demand growth in all key markets. Both SPRYCEL and ORENCIA were strong, with sales increasing 22% and 18% respectively. Certainly, the highlight in our virology business was the launch of our Dual Regimen in Japan and Daklinza in Europe for hepatitis C. We reported…

Thanks, Charlie. And I just like to remind everybody that in addition to Lamberto and Charlie, both Giovanni and Francis are here to handle any questions you might have. Travis?

[Operator Instructions] We'll take our first question from Seamus Fernandez with Leerink.

So just a question as it specifically relates to thinking about kind of metrics in third-line lung cancer. Just hoping that you could give us a little bit of a rundown of Bristol's thoughts on the information that you have in terms of the 1-year survival metrics, 2-year survival metrics for the squamous non-small cell lung cancer patient population. And also as it relates to the CheckMate -017 study, any update on the timing of the readout and any sort of -- is it possible that the events are coming in more slowly and that this could slip out? Or do we still expect CheckMate -017, the interim look, to occur this quarter?

Good morning, Seamus. Thank you. It's Francis. First of all, let me take the -063 study. We're looking forward to presenting that data next week at the Thoracic Oncology meeting. The -- as you know, it's very difficult to compare across trials in this area because -063 is a highly pre-treated, third-plus line of therapy with no really other approved options or recommendations. But as we've looked at this roughly similar patient population, I would suggest the historical objective response rates in the low to mid-single digits. And you may recall, at ASCO this year, we presented an analysis of SEER registry of Medicare data that showed that perhaps less than 20%, in fact 18%, of the third-line squamous non-small cell lung cancer patients were actually alive after a year. Let me move on to CheckMate -017. As you commented, the timing of the analysis on this study is event-driven. And that means that patients -- the patient mortality will determine when we do the analysis. As we've seen in previous I-O studies, the mortality rate continues to slow down. Our present projection -- and of course, this is a projection, but our present projection is that we will trigger the -- or the Data Management Committee will trigger an analysis at the -- towards the end of the year. And I think it's just worth commenting that, of course, there will be some time, some weeks while they do their analysis. So we remain on track to -- with the present projections to have a year-end triggering of that analysis.

Our next question comes from Greg Gilbert with Deutsche Bank.

First, 2 quick ones for Charlie. Can you just -- sorry if we missed this, but the other income line, can you break that up between the interest and other stuff? And then on revenue guidance, is Daklinza the offset to the BARACLUDE generic hit that I'm pretty sure wasn't in your guidance previously? And then lastly, I just had a bigger-picture question about the anti-PD-1 space. Are there any observations you can make about Merck's go-to-market strategy so far that sort of shaped how you're thinking about your Opdivo go-to-market strategy?

Giovanni, why don't you take the Merck question first and then Giovanni -- and Charlie will continue with the rest.

Yes. Good morning. This is Giovanni. With respect to the question on go-to-market strategy, obviously, our go-to-market strategy for Opdivo will be informed primarily by the very large experience we have with Yervoy and the fact that we have developed a very innovative customer model. When we launched the Yervoy metastatic melanoma, we've developed a very significant experience in terms of the ability to educate physicians in immuno-oncology. So we are obviously looking at building on the very significant experience that we have with Yervoy and further developing our go-to-market strategy with Opdivo when we get approved. I'll just remind you that as a result of, clearly, the strength of the data, the survival advantage of Yervoy and the strategy we followed to educate physicians in the academic setting and the community for Yervoy, we continue to see very, very strong trends in both settings. It's really too early for us to comment on the launch of Keytruda, but as Charlie said, our recent weekly trends in terms of sales continue to be very strong.

Yes. I mean, just quickly on -- Greg and you can follow up with John and his team on the details of the other income. Our interest income expense in the U.S., that specifically is relatively the same as it was in the third quarter of last year. And as we've commented before, the transaction services agreement that we have with AstraZeneca pretty much just starts to wind down significantly by the end of the third quarter. So we really won't see much further income from that TSA income on a go-forward basis.

The next question comes from Jami Rubin with Goldman Sachs.

Just a couple of questions. First, Francis, for you. Are you still committed to initiating a Phase III combo lung study, Yervoy, nivo and lung? And at some point, will you show us the data from that -- I guess, it's your -012 trial that I would expect you would use to determine which combination or which strategy to go forward with that Phase III trial? And then secondly, do you think that -- is there a read across from the third-line squamous to the second-line trial that hopefully reads out by the end of this year? And if you could kind of confirm whether or not we will actually see overall survival with the third-line squamous study next week.

Good morning, Jami. Thank you. So first of all, let me talk about the -012 study, which as you recall is a multi-arm study looking at different combinations in lung cancer. It's still ongoing. It's still producing clinical data that's informing our design of our Phase III study in combination. We have sufficient information to complete the design and we're heading towards completing that by the end of the year. I can't actually comment at this point exactly on the design, but you will be able to read about it in clinicaltrials.gov when it gets posted probably at the beginning of next year. As far as the -012 data, we've not yet decided the venue or the timing of the 1-plus-1 data, for instance, although you may see some of the data from the Avastin, nivo study at the Thoracic Oncology Study -- Symposium next week. As far as the read-through between the third line and second line, I think I would say that it is difficult to compare across different patient populations. To repeat, the third-line Study -063 actually includes fourth-line and fifth-line patients, patients who are actually progressing as they went into the study. The second line is a less severe population, although it does really include all comers both in terms of PD-L1 positive and negative. So overall, as we get more data, we remain confident in the Yervoy, Opdivo lung combinations and we're looking forward to initiating the study. Thank you.

Our next question comes from Tim Anderson with Sanford Bernstein. Timothy Anderson - Sanford C. Bernstein & Co., LLC., Research Division: A few questions, please. Just hoping to clarify on the interim look. So on the interim or any interim looks that will be done on -017 and -057, will you have access to what the data shows? Or will you otherwise be informed in any meaningful way of those findings, apart from if there were to be early stoppage? Or is that something that only the DMC and potentially FDA looks at? Second question is I think you and your competitors monitor each other's moves very closely in the PD-1 space. I'm hoping you can give us your thoughts on whether Roche and Merck might be able to file their own products for approval in lung in early '15, which seems like it could be possible. And then just a clarification on the nivo plus Yervoy combo trial, should we assume that, that will be in PD-L1 negative patients or in all comers?

Tim, good morning. So let me just talk first throughout the interim data look. A trigger for this is a predetermined number of events, as I've already mentioned, that is slowing down. This is something we've seen before with Yervoy. It's, of course, good for patients and it will trigger the Data Monitoring Committee to do the analysis. We -- even though this is an open study, it's blind to us. We have no access to the data unless it hits the stopping rules, as you say. As far as the moves of our competitors, I'm totally focused on our program and I clearly can't really comment on what they may do or what they may not do. And as far as the combination time line, I would just reiterate that we're looking to have that -- in this, the study design and preparation completed by the end of the year and you will be able to read the fine details of that when it's posted to clinicaltrials.gov presumably in that first quarter. Thank you.

The next question comes from Chris Schott with JPMorgan. Christopher T. Schott - JP Morgan Chase & Co, Research Division: I just had 2 here. First, you've expressed some excitement about the opportunity for PD-1s in hematologic cancers and whether you're seeing some initial data at ASH. So I look at these markets, we're seeing a rapidly evolving standard of care. Can you just talk a little bit about where you see the PD-1s ultimately fitting into treatment paradigms here? And the second question was on tax. The company has always done a great job and has one of the lowest tax rates in the group. In light of all just the recent tax scrutiny that's been out there, can you just talk about how you're thinking about your tax rate over the next several years? I know that those double Irish stuff, et cetera, may not apply. But just generally speaking, how do you think about your tax rate over time?

Charlie. And not an expected question on taxes.

Okay. Backing up, thank you, Chris. As you know, we're not giving any guidance at this point on any particular line item. We're in the middle our budget process and we'll talk about our '15 tax rate in January in the normal cycle. I would say, overall, and I've said this before, that our tax rate is primarily influenced by earnings mix, but also there are onetime items and also very importantly, our tax planning opportunities which allowed us, as you pointed out, to have a fairly good tax rate compared to our peers. And these tax planning opportunities are something that we will continually assess. So until we get to January, I don't think there's more perspective I can provide you on the tax rate.

You were talking about the Irish tax double.

Yes, I mean, we don't have anything related to the Irish double tax.

Good morning, Chris. Let me first say that we were very happy to receive Breakthrough Designation for Hodgkin's lymphoma from the FDA and are looking forward to presenting that data and the non-Hodgkin's data at ASH in December. This is clearly in a late-stage group of patients and it's -- you'll have to wait and see what the data is. But I think there is good rationale as well as the data we have to start out at the late stage, but I think it will be potentially attractive because of the potential for long-term survival and the underlying biology of hematological tumors. Thank you.

Our next question comes from Andrew Baum with Citi.

3 questions, please. Firstly for Francis, regarding the -017 trial, should I assume that the early stopping rule for the interim has a similar p-value or confidence interval compared to that of the final analysis? It obviously depends which stopping rule you are using. Second, could you remind us when we should expect the first CD137 data? And then finally, could you update us, please, on the filing strategy and timing for Yervoy in the adjuvant melanoma setting?

Good morning, Andrew. So I'm not going to get into the detail of the stopping rules either for the interim or the final analysis. As far as the 137 Urelumab data is concerned, those trials are ongoing and we will -- haven't yet decided when and where we might present that data. Certainly, we're considering, for competitive reasons, perhaps not being as forthcoming as we might have done in the past. And let me just move to our plans for the adjuvant melanoma filing. As you know, the data we presented at ASCO in the -029 study was actually quite strong and we are having productive discussions with the FDA and plan to file the adjuvant melanoma data for regulatory approval. As we traditionally do, we'll provide further information when it's accepted by the FDA. But we believe that adjuvant melanoma is a significant area of unmet need and is an important part of our strategy in melanoma.

Yes. And this is Giovanni. And obviously, from a commercial perspective, adjuvant melanoma is treated in many of the same academic institutions where we have established a really strong presence through Yervoy. As Francis mentioned, it's an area of very high unmet medical need. And as we think about preparing for a potential launch, obviously, our focus would be on subpopulations of patients at high risk where we believe that Yervoy can have a very, very meaningful role to play.

Our next question comes from Mark Schoenebaum with the ISI Group. [Technical Difficulty]

Wow. That's a first.

Sorry, guys. Can you hear me?

Yes.

We got you, Mark.

Okay. I apologize for that, John. Francis, thanks for all the disclosures. This call has been clear. I really -- I just really appreciate that. If I may ask just one more question on the -017 trial, can you just clarify, I wasn't sure from your remarks, it sounds like you'll have the data in-house, interim data in-house, by the end of the year. Are you still certain that you'll have -- you'll be able to disclose something to investors by the end of the year? And if indeed the trial continues, due to the fact that this is an interim analysis, can you help us out, when do you believe roughly the final analysis would be triggered, please? And then actually, one question for Lamberto, if I might. Just the M&A environment has heated up and cooled down, et cetera. Can you just update us on Bristol's willingness to think about doing an M&A transaction that might be on the larger side? So something that's clearly not defined as a tuck-in.

Good morning, Mark. So thank you for that question. And I do want to clarify this. We will not have the data in-house unless the study stops. So this is -- this data analysis is triggered by the event and it's performed by the Data Monitoring Committee. It looks like from our projections and of course, they're still a projection because the mortality rate is slowing down, that we will trigger that analysis about the end of the year. Then of course, there's the time for the analysis beyond that. So there will not be a disclosure when the analysis is triggered, it will be if and when the study has stopped, when the analysis is done. And as far as the final analysis is concerned, I mean, it becomes even less certain in a way because obviously, it's still an event-driven study. So we would imagine towards the end of next year, but that is a very distant projection. Thank you.

So about M&As. It is right. There's a lot going on outside there and our point of view has not changed. We continue to look at opportunities of acquiring products, technologies, pipelines in areas that are significant and key to us. We do not consider -- we have never considered inversions that -- as a goal for us because, as I said, we are interested in products and technologies. In terms of size, we are looking at opportunities of a different size. And we believe that we have the financial capacity to go to a large-size operation if we find the opportunity worth investigating. But we are also continuing to look at smaller opportunities and we have announced some and we will announce some more in the near future. Charlie, you want to add something here?

No, I mean, I would only add that if you look at our track record, we haven't really done anything of size, but we've always said we are agnostic to the type of deal, structure of deal and size of deal, I mean, just to conclude with that.

Our next question comes from Vamil Divan with Crédit Suisse. Ronak H. Shah - Crédit Suisse AG, Research Division: This is a Ronak filling in for Vamil. 2 quick questions. One is on the patent side. Can you give us an update on your litigation against Merck related to Opdivo? And what are the relevant milestones we should be watching then in the time frame for resolution? And then second, just to get a better sense of your strategy in first-line melanoma, have you seen the results of CheckMate -069 at this point and have had any discussions with regulators on that data? And if not, how do you plan to leverage that trial moving forward?

Very quickly on the IP, we believe in the strength of our IP's position. We will continue to defend it and there is nothing new to announce here. We made it clear that we are not going to block anything that could be negative for patients, but obviously, we are going to defend our intellectual property rights and we are going to do it here and in Europe and elsewhere.

Ronak, good morning. We do have the data analysis in-house for the -069 study and I'm encouraged by the data. We are working with investigators to determine the venue and timing of the presentation and publication. But I would say that it's premature to disclose any details of our regulatory strategy at this time. Thank you.

Our next question comes from Dave Risinger with Morgan Stanley.

So I have a couple of questions. First, what I have down on the time line for the -057 non-squamous Phase III trial is that the completion is estimated on clinicaltrials.gov to occur next summer, yet, you were, I think, speaking previously about an interim look early next year. So my first question is could you update us on those 2 time line events? Second, can you explain why the interim look is so close to the conclusion of the trial? And then my second question is a little bit longer -- or my next question is a little bit longer one and it's about Yervoy. So there are sort of 2 views about the outlook for Yervoy in the near term. One view from an expert that we had hosted is that Yervoy use in the community may step up because patients have to fail Yervoy before going on to Keytruda and thus, maybe the use of Yervoy could step up in the near term. A different view is that the threshold to stop Yervoy may be lower than before for some patients because there's obviously an alternative. And so for a patient who is struggling with the side effects of Yervoy, they can drop the therapy at the first sign of toxicity and switch to Keytruda so Yervoy sales could decline in the near term. Could you please give us your view on which scenario is more likely over the next few quarters?

Good morning, Dave. So let me clarify exactly what's going with -057. It's very similar to -017 insofar as it's an event-driven study and that the event rate is slowing down, perhaps it's slowing down even more than -017. So there is an anomaly there. As I said, the -- we will -- we're projecting an interim analysis in the first half of next year and of course, it would not be rational to expect a final analysis in the middle of the year according to clinicaltrials.gov. So it's -- like -017, the final analysis will also project out further as well.

Yes, Dave. This is Giovanni. With respect to your question on Yervoy melanoma, obviously, this is a very dynamic marketplace with some events happening and we believe significant developments going forward. I would start from our current position. So when you look at Yervoy today, we have very strong trends. Over 60% of our business in the community is in the community and that is the most rapidly growing segment today. 70% of our use is in the first-line setting. And so the first point is that, clearly, the role of Yervoy in metastatic melanoma is very well established across first and second line and in both the academic setting and the community setting. We are also, I must say, seeing not only very high penetration in BRAF wild-type, but we've also seen a very good increase in the use of Yervoy in the mutant segment in the last couple of quarters. We've always said that it is possible that in the short term, there is volatility in the use of Yervoy. But remember, that the -- that Keytruda is approved in a very specific patient population, in patients that have failed Yervoy and that was confirmed recently in the latest update of the NCCN guidelines. And obviously, we are very focused on the medium and the long term. And there, we continue to be very focused on the fact that combination use in the first-line setting can provide very, very strong hope for a patient's survival in the first year after the therapy started. So hopefully, that adds a little bit on our perspective there.

Our next question comes from Alex Arfaei with BMO Capital Markets.

I have 3, if I may, first 2 for Charlie. How should we think about royalties for nivo in Japan? I see it's registering on your financials this quarter. Could this be a meaningful opportunity for you? And then looking ahead, how should we think about your margins longer term? Obviously, Eliquis will be a drag, given the partnership structure, but your product mix tends to be shifting towards more higher-margin products. And as immuno-oncology becomes a larger contributor, is it fair to expect significant margin expansion opportunity? And then the last one for Francis. Based on everything you've seen so far, is there anything to suggest that more advanced lung cancer patients are more likely to be PD-L1 positive because of prior treatments? And just to clarify, will we have 2-year OS data from the -063 study?

Yes. So thank you, Alex. So on the Ono collaboration, as we've disclosed in the last quarter, we are partnering -- Ono had the rights to nivo in Japan, South Korea and Taiwan and we partnered with them with 3 of our earlier-stage compounds to form an alliance, whereby, where it's in combination, we will share equally between the 2 companies and where that it's used in as a single agent, that the lion's share will go to each partner in which they share the product. So that's the Ono collaboration. As it relates to gross margin going forward, as you point out, we have Eliquis, which is below 50% margin based upon the deal that we have with Pfizer. And the rest of our products are slightly higher than our overall corporate margin. So you have to really look at the overall mix of products going forward.

So Alex, let me just comment that we don't have the PD-L1 status in this patient population. There isn't any particular reason to consider they'd be more likely or less to be PD-L1 positive or not. Certainly, we've seen in other studies that we seem to have that representational proportion of PD-L1 positive patients that are the same and more or less across them. As far as the landmark data in terms of 1-year or 2-year, you will see in the abstract and it's presented, published next week, sixth-month data, but the presentation will be on the 1-month data, so...

1 year.

I'm sorry. The quantitative data. So 6 months in the abstract, 12 months in the presentation. Thank you.

Our next question comes from Steve Scala with Cowen and Company.

I have 3 questions. There is an expectation that Bristol's long-term multiyear tax rate will be high teens. In fact, I think Bristol may have actually said that at one point. If not, please correct me. But has that expectation changed? Second, to hit the top end of Bristol's 2014 guidance range, Q4 EPS needs to be only $0.42, that's down 18%, making it by far the worst quarter of 2014. Other than BARACLUDE, which also impacted Q3, I'm not sure why this would be the case. So can you amplify? And then the third question. Bristol competitors talk a lot about OX-40 and GITR, yet Bristol doesn't have any at least later-stage programs in these areas. Is that because Bristol simply hasn't come upon a good molecule? Is it due to IP? Or does Bristol question the value of the target? And if it's the latter, why is Bristol skeptical?

Thanks, Steve. Charlie. So I'll answer the first 2 parts of your question. So in regard to tax, yes, we previously have stated that we expect the tax rate, not indefinitely, but over the medium term, to be in the high teens. Clearly, that's a function of the factors that I mentioned before, which is earnings mix. And we did see a little bit of up-skew towards higher product earnings in the higher tax rate areas, so that's impacted it slightly. The Diabetes business going away actually hurt the rate a little bit as well. But on average, looking at all the different factors that I mentioned, including our tax planning strategies, yes, I still feel in the medium term that that's a good barometer to use. As it relates to trends, vis-à-vis, how we are looking through the 3 quarters and then in the fourth quarter, a few things to recognize. The bullets of the BARACLUDE impact is going to happen in the fourth quarter, where also as I mentioned earlier, the Diabetes royalties, the way they're structured and the -- so we get a pretty good royalty on the first $500 million. We really don't get much as we get to the balance of this year. We also have the transaction services agreement with AstraZeneca, whereby that pretty much drops off in the fourth quarter. And then we do have commercial investments that we are continuing to make in the business around Yervoy as we begin the preparation for the launch of Opdivo, Eliquis DVT. And also as you look at the R&D spend through the 3 quarters and what we see in the back half of the year, we do see, particularly in regard to I-O, expanded patient enrollment continuation of new study starts, including several clinical collaborations exploring combination regimens. And then we do have some costs related to comparator-arm studies. So by and large, we do see an increase in our overall expenses from a trend-wise in the fourth quarter.

Steve, good morning. Let me just reemphasize that we totally believe that combinations of I-O agents will play a critical role in bringing this potential long-term survival to the broadest number of patients. And of course, you've seen that in terms of our Phase III starts and potentially data over the next 12 to 18 months. We've also got early development, 3 early development combinations ongoing and our last portfolio in discovery and actually in preclinical development. And I think, notably this year, we've -- as broad as that is, we have actually supplemented that with some collaborations, most notably the Five Prime, the Dana Farber. So yes, we are committed to bringing forward a broad portfolio of potential immuno-oncology combinations. Thank you.

Our next question comes from Colin Bristow with Bank of America Merrill Lynch.

Maybe switch gears to hep C. Can you talk about how you see your triple regimen being competitively positioned, especially in line with treatment landscape that's moving towards shorter-duration regimens? And then as you're assessing the market opportunity, to what extent do you see price being a competitive lever for those competitors who lack clinical differentiation?

So thanks, Colin. We're reviewing the recent emerging clinical data from our 2 global 3DAA studies, which you recall is UNITY 1 and UNITY 2 and that those will be presented as soon as the upcoming AASLD the meeting and we'll be evaluating our regulatory strategy in the U.S. as well as other countries in the context of this very fast-moving environment.

Yes. I would say to answer -- this is Giovanni. Good morning. To answer your question on our strategy and particularly with respect to pricing, a couple of comments there. The first one is from a strategic perspective, our priority remains to identify areas of unmet medical needs, of high unmet medical needs and develop regimens that address those. And as we've discussed before, obviously, that is different, depending on the region and the market. With respect to pricing, I -- our belief is that the value continues to be the primary driver in this market like in other markets. And so we are very focused on delivering regimens that deliver high value from a clinical perspective. But obviously, we are working in Europe, in Japan, in other geographies with payers to ensure we work in partner with them to ensure that our products are reimbursed.

The last question today will come from Tony Butler with Guggenheim Partners.

Francis, just one somewhat broad question again on Opdivo. Certainly, at ESMO, there was a great deal of discussion around the theme of PD-L1 positivity, but more importantly, around enriching populations or trials related to PD-L1 positivity. And as you think about the combination trial, which you said would be announced at least early next year in Clinical Trials, but also in other data sets, what's your view around enriching for PD-L1 positivity? Or do you dismiss that theme totally?

Tony, good morning. Our approach to this has been -- and I think let me just zoom out a little bit to our biomarker strategy. The -- our entire development program has incorporated biomarker testing, as you know. And we want to understand better not just the objective response rate, but also the linkage of PD-L1 positivity to the overall survival. So all our patients have tumors tested and PD-L1 positivity measured. It's clear that in some cases, PD-L1 -- there is a better response rate with PD-L1 positivity, but also both in monotherapy and particularly in combination therapy, we've seen that patients who are PD-L1 negative do get responses. And we do know that those patients, the PD-L1 negative patients, do contribute to overall survival. So our approach to this is to collect the data, to do the study and we'll look at the data and put together our regulatory submissions based on what that data shows. Thank you.

Thanks, Tony. And thanks, everybody, for your questions. Lamberto, any closing remarks?

Thank you very much, everybody. We had a very good third quarter, good in terms of financial results as well as clinical results and regulatory milestones and we are working actively at delivering good to further in 2014 and beyond. Thank you.

That concludes today's presentation. Thank you for your participation.