福泰製藥 (VRTX) 2018 Q3 法說會逐字稿

Welcome to the Vertex Third Quarter 2018 Conference Call. This is Michael Partridge, Senior Vice President of Investor Relations for Vertex. Tonight, we will review our financial results, our continued progress to develop new medicines for all people with cystic fibrosis and recent advances in our research and development pipeline. Dr. Jeff Leiden, Chairman and CEO; and Ian Smith, Chief Operating Officer, will provide prepared remarks this evening. Stuart Arbuckle, Chief Commercial Officer; and Dr. Reshma Kewalramani, Chief Medical Officer, will join us for Q&A. We recommend that you access the webcast slides on our website as you listen to this call. This conference call is being recorded. A replay will be available on our website.

We will make forward-looking statements on this call that are subject to the risks and uncertainties discussed in detail in today's press release and our filings with the Securities and Exchange Commission. These statements, including, without limitation, those regarding Vertex's marketed CF medicines, the ongoing development and potential commercialization of our triple combination regimens for cystic fibrosis, Vertex's other programs and Vertex's future financial performance are based on management's current assumptions. Actual outcomes and events could differ materially.

I will now turn the call over to Dr. Jeff Leiden.

Thanks, Michael. Good evening, everyone. I'm pleased to say that we have made tremendous progress across our business in 2018 and are well on track to achieve the key goals that we outlined to start this year. In CF, approximately half of all patients are now eligible for a Vertex CF medicine, and we're progressing rapidly toward the development of a single medicine to treat any CF patient with one or more F508del mutations, approximately 90% of all people with the disease.

In our development pipeline, we have initiated the first clinical study of CTX001 in beta-thalassemia with our partner CRISPR Therapeutics and are progressing VX-150 for the treatment of pain. And in research, we're preparing to advance the first of our potential medicines for alpha-1 antitrypsin deficiency into clinical development by year-end and to move other molecules into development in 2019.

First, to our recent progress in cystic fibrosis. CF is a progressive multi-organ disease that is present from birth and a key part of our strategy is to treat patients as early as possible to slow or prevent the progression of disease. To that end, we gained important approvals throughout 2018 that now allow us to treat eligible children as young as 1 year of age with KALYDECO and as young as 2 years of age with ORKAMBI.

With SYMDEKO, demand continues to be strong following the U.S. approval in February. The launch of this medicine in the U.S. has been a significant driver of revenue growth in 2018 as new patients initiate treatment. And we also plan to submit a supplemental New Drug Application to the FDA later this year to gain approval in eligible patients as young as 6 years of age.

In Europe, we anticipate approval of this medicine before the end of this year, which will further increase the number of patients eligible to be treated with the medicine for the underlying cause of their disease.

We've also reached multiple important reimbursement agreements across the world to provide access to ORKAMBI, enabling patients in many countries to treat the underlying cause of the disease for the first time. Many of these agreements, including those recently signed in Denmark and Australia also provide a pathway to access and rapid reimbursement for certain future CF medicines. Importantly, these portfolio agreements provide certainty to patients that they will have immediate access to future innovations in CF from Vertex. We remain in discussions regarding this type of agreement in other countries that have yet to reimburse ORKAMBI.

We're also rapidly progressing our 2 Phase III programs, evaluating 2 different triple combination regimens that include a next-generation corrector, either VX-659 or VX-445. In September, we completed the enrollment of the 2 Phase III studies for the VX-659 triple combination regimen and expect data from these studies in late 2018. We expect to complete enrollment for the VX-445 Phase III studies this quarter and to report data from these studies in the first quarter of 2019.

We will evaluate data from both programs with the goal of choosing the best regimen to submit for regulatory approval. We remain on track to submit a New Drug Application for a triple combination regimen no later than mid-2019. In parallel, we've also recently initiated studies with both of our triple combination regimens in patients ages 6 to 11 years of age with the goal of gaining approval for this group of children as quickly as possible.

We also continue to innovate in CF to develop even better regimens for the future, including potential once daily triple combination regimens with our potentiator, VX-561, in regimens that contain other next-generation correctors that may have enhanced profiles. We expect to advance one or more of these novel, next-generation correctors into early clinical development in the coming months.

Beyond CF, our research strategy is well-defined and focused on the development of transformational medicines for serious specialty diseases with large unmet medical need. And we're advancing a portfolio of such potential new medicines into the clinic. The diseases we are targeting all have well understood biology where we can use or create early clinical markers that support the potential for transformative benefit and rapid development time lines.

We are establishing proprietary, scientific insights into a number of these diseases, just as we did in CF and CFTR. And we are bringing significant resources to bear to both advance our development-stage assets and also to bring other potential new medicines from research into early clinical studies.

With our partner CRISPR Therapeutics, we're advancing CTX001 as the first gene editing treatment for both sickle cell disease and beta-thalassemia using the CRISPR-Cas9 technology. This is a cutting-edge, transformational technology that holds great promise in the treatment of many diseases, and we're pleased with the rapid progress we are making together with CRISPR toward dosing the first patient with a CRISPR-Cas9 gene editing therapy.

Our pain program is another example of Vertex's innovative science translating to the clinic, where Phase II data generated to date for VX-150 provided the first clinical validation for the sodium channel 1.8 mechanism in the treatment of acute and chronic pain. We expect to have Phase II results for VX-150 in the third type of pain, neuropathic pain, in early 2019.

We now also plan to initiate a Phase IIb dose-ranging study of VX-150 in acute pain following bunionectomy surgery that has the potential to support future pivotal development of this medicine in acute pain. Acute pain represents a multibillion-dollar opportunity where many of the conditions that lead to acute pain are treated by physicians that can be reached with a specialty sales force.

Given the significant need for highly efficacious pain medicines that do not have the addiction, tolerability and abuse issues associated with opioids, acute pain is an opportunity that is well-suited to inhibitors of the NaV1.8 sodium channel, and we're also continuing to invest in the discovery of other potential pain molecules, including additional NaV1.8 inhibitors as well as medicines targeting other new mechanisms.

We're also advancing a portfolio of small molecule correctors for the treatment of alpha-1 antitrypsin deficiency or AAT. And I'm pleased to report that we expect to move the first of these compounds into clinical development later this year. There are many similarities between AAT and CF, both in the biology of the disease and how we may target its cause and in the potential to establish early clinical proof of concept and rapid development pathways for new AAT medicines.

Like CF, AAT is a protein folding disorder caused by mutations in a single gene where more than 90% of patients have one common mutation that results in life-shortening, systemic complications primarily in the lung, but also in the liver. We believe our scientific expertise with CFTR protein folding, coupled with our development experience, position us well to rapidly advance our novel small molecule approach to the treatment of AAT.

I look forward to updating you on this program as we enter development later this year and to providing additional insight into our research programs, including those in focal segmental glomerulosclerosis and others as they progress in 2019.

As we move our internal R&D pipeline forward, we're also continuously evaluating external opportunities where our research and business development strategies are fully aligned. With our strong financial profile, we have increased ability to in-license or acquire assets or establish scientific collaborations that provide access to unique product opportunities and technology platforms.

We've completed a number of such transactions in recent years, and we've expanded our internal team devoted to finding and evaluating potential opportunities for scientific innovation with the goal of further broadening our pipeline and scientific expertise in the future.

We believe that it is a combination of our R&D strategy with our business model that truly differentiates Vertex as a biotechnology company with the potential for highly profitable, long-term growth based on a portfolio of transformative medicines for multiple serious diseases. This year has been marked by significant progress across research and development and in our efforts to bring our CF medicines to more patients globally, which has led to significant growth in our revenues and earnings, which Ian will now discuss in more detail.

Thanks, Jeff, and good evening to everyone. I'm pleased to review with you our third quarter 2018 financial results and to discuss our outlook for future financial growth as we approach 2019.

Revenues first. Total CF product revenues of $783 million in the third quarter of 2018 represent a 42% increase compared to the $550 million we recorded in the third quarter of 2017. We continue to see significant revenue growth as we increase the number of patients treated with our medicines globally.

The third quarter included $255 million in revenues from the launch of SYMDEKO in the U.S., which is the primary driver of the rapid growth in our total CF revenues. Demand for SYMDEKO remains strong, and we continue to receive positive feedback from patients and physicians and strong coverage for SYMDEKO across both public and private payers.

Similar to prior quarters, demand for SYMDEKO has come from all groups of eligible patients: F508del homozygous patients initiating treatment for the first time, patients who discontinued ORKAMBI coming back to initiate therapy with SYMDEKO and also patients switching from ORKAMBI to SYMDEKO.

Our third quarter 2018 non-GAAP combined R&D and SG&A expenses were $379 million compared to $334 million in the third quarter 2017. This increase was primarily due to the advancement of our portfolio of triple combination regimens for CF and the investment to support the treatment of patients with our medicines globally.

Non-GAAP net income for the third quarter of 2018 was $282 million with an earnings per share of $1.09, compared to non-GAAP net income of $136 million with an earnings per share of $0.53 for the third quarter 2017. Our non-GAAP net income and EPS has more than doubled compared to last year largely driven by strong growth in the total CF product revenues.

We ended the quarter with approximately $3.1 billion in cash, cash equivalents and marketable securities compared to $2.1 billion at the beginning of this year.

Today, we also reiterated our guidance for total CF product revenues of $2.9 billion to $3 billion and for combined non-GAAP R&D and SG&A expenses of $1.5 billion to $1.55 billion.

I will close with a few comments on our expectations for future financial growth and how we anticipate providing revenue guidance next year as well as how we may start to record noncash tax charges. We have already created a strong financial profile for our business, and we anticipate that our revenues will continue to grow next year.

In 2019, revenue growth will be based largely on the impact of SYMDEKO and SYMKEVI launches, the impact of label expansions across our CF portfolio and the recent completion of multiple reimbursement agreements. The timing of further revenue growth from countries where we do not currently have reimbursement for ORKAMBI is unpredictable. Therefore, when we provide 2019 guidance for total CF revenues early next year, it will be based only on the geographies where we have established reimbursement agreements at that date. If we gain additional reimbursement agreements in 2019, we will update you on our guidance as appropriate.

This is consistent with how we provided guidance at the beginning of 2018. Beyond 2019, continued revenue growth will be driven largely by the potential approval and launch of a triple combination medicine for the large group of patients with one minimal function mutation and one F508del mutation who are not eligible for our currently approved medicines.

Now to taxes. I would note that in 2019, we may begin recording a noncash tax provision at an effective tax rate in the low to mid-20s as a result of the improving profitability of our business and our future strong financial outlook. The vast majority of our tax provision would be a noncash expense until we fully utilize our net operating losses. This will be further discussed in our 10-Q filing.

The financial profile of our business is strong, and the current year will be important period to define our future growth as we obtain Phase III data from our triple combination regimens for CF, provide access to each of our CF medicines to more patients globally, progress multiple new medicines in development and also to advance new approaches for the treatment of serious diseases from research into development. I look forward to updating you on our progress.

With that, I will open the line to questions.

(Operator Instructions) And our first question is from Phil Nadeau from Cowen and Company.

Two questions for me, one on competition and one on kind of the financial guidance or financial comments that you just made. Ian, in your comments, you talked about 2019 sources of revenue growth. Can you specifically talk about the revenue growth outside the U.S., which I know you said was unpredictable, which countries could be drivers of growth where you recently got expansions, which ones are possible for next year and which ones are total wildcards?

Yes. Thanks for the question Phil, and thanks for the comment on the good quarter. I do want to kind of replay some of the comments I just made, which is when we think about 2019, first of all, we see 2019 as being the continued growth year for our revenues. When we looked over the previous years and we look over future years, I think what we're seeing here is, we're treating more patients. We're treating more patients globally. And therefore, that's driving our revenue line year-on-year and into the future. And so it's a really nice profile. As we think about more proximally in time and we think about going from 2018 to 2019, we should continue to see growth even without reimbursement outside the U.S. And that growth would be driven by the continued launch of SYMDEKO in the U.S. and getting a full year's contribution for patients that are on drug. But then also outside of the U.S., SYMDEKO, which is known as SYMKEVI outside the U.S., is also launching in certain markets where it is already reimbursed. So we see baseline growth from already approved geographies. More specifically to your question of where are the geographies that we could anticipate will possibly attain reimbursement approval, to have an impact in terms of growth on our revenues, it would have to be in those larger markets. And principally, those would be either the U.K., France or Canada. I do want to replay though, this is unpredictable in terms of timing. And in particular, in the U.K., it is the second-largest CF market. If we are to obtain reimbursement over there, then we are launching the drug from anew. So we're in a launch curve in the U.K. It's not that we're just switching on a large bolus of revenue. We have to launch the drugs in the U.K. We continue to make good progress with discussions with all 3 of those countries. However, resolution, whether it is in 2019 or whether it's in 2020, we'll update you as we progress.

That's really helpful. And then second on competition. After the close today, Galapagos and AbbVie had a couple of announcements, some new data as well as a change to the collaboration. And Proteostasis recently announced some data for their programs. I'm curious to get the team's updated thoughts on competition from those 2 parties, how real is it? What do you see as the strengths and weaknesses?

Yes. So this is Jeff. Thanks for the question. As you know, we don't typically comment on individual competitor data. I think actually the data sort of speaks for itself. But what I would like to comment on is sort of how we view our competitive position and our data and sort of where we are in the process. I think we've learned a lot over the last year. In fact, I think our competitive position has improved over the last year significantly. One of the things we've learned is that, it's going to be a world of triples. And it's going to be a world of triples pretty quickly. And the reason for that is that in order to really get to all patients, particularly the het/min patients who are so difficult to treat and to maximize benefit for the homozygous patients, it's pretty clear now that you need a triple. Any company needs a triple. And that doesn't only mean you have 3 components. It means you have 3 medicines that can be combined, co-formulated, that have the right PK, that can provide really superior efficacy. And I think the bar in efficacy has gone up very significantly, we'll talk about that in a minute. And also importantly is that they're tolerable and safe. These are medicines that kids are going to take for the rest of their lives. And we've learned a lot about our regimens in that regard in the last year. We know that we have 2 different triple regimens that have now been in hundreds of patients. They are co-formulatable. They do have the right PK. They raise the bar in efficacy very significantly into the double digits for the het/min patients and into the mid-double digits for the homozygous patients. That's going to be the bar, I think, for the future. And most importantly to us, we're well into our Phase III trials, which are enrolling ahead of schedule to the point where we're very confident we'll be able to file one of those regimens no later than mid next year, which would allow us to launch those drugs and get patients on them pretty quickly in the coming years. And so I think from our perspective, we like where we are. We have multiple options. And most importantly, we have clinical data both on efficacy and on safety with these drugs, which is a high bar, frankly.

Our next question is from Geoff Meacham from Barclays.

Just have a couple. The first one is on the triple selection process just for the NDA. When you guys think about it, you likely won't have much in the way of mature data on exacerbations or BMI when you pick the regimen. And that may matter actually over the long term. So is it just about safety, tolerability and FEV1? Or are there other parts of the profile that you think will be impactful that you're looking for?

Yes, Geoff, this is Jeff Leiden. You're right. I mean, I think at the point we can pick these triples, the primary data that we'll base that on is FEV1 data and safety and tolerability data as well as the whole profile of these drugs. Remember, we have a lot of preclinical and early clinical data as well. I think the good news is there hasn't been enormous disparity between many of those long-term outcomes and short-term outcomes. And what I mean by that is that, at least in our experience so far, drugs that have a very positive, acute effect on FEV1 and a lowering of sweat chloride that's significant acutely have also turned out to have these very profound long-term effects. And in fact, as you should know, one of the interesting things is that KALYDECO and say, ORKAMBI, which have a bit different acute profiles, ORKAMBI being less potent acutely, actually have very similar long term profiles, which suggest that if you can really hit CFTR effectively in these patients, you will get the long-term benefit. So I think the risk of a disconnect between an acute effect and a long-term effect is rather small. Now obviously, we will have the long-term effects of both of these drugs within a relatively short period of time. But you're right, probably not by the time we decide which one to file.

Okay. Just a follow-up for Ian, and I know you haven't given a 2019 outlook, but when you plan for OpEx spend, you'll have the triple launch pending the data obviously and then a bunch of new programs in R&D. So how are you guys thinking about that in the context of also showing some operating leverage?

Geoff, thanks for the question. I really appreciate you pointing out that we've got this continued investment and commitment to CF. I just want to cover that because we do sometimes get asked the question whether when we deliver on these Phase III trials, are we declining our investments in CF? And I actually want to maintain the position that we're not. We continue to invest in CF. In fact, those patients that were in our Phase III studies roll over to open-label studies. Those open-label studies we had found to be incredibly important because we got the longitudinal data to show the benefit of these medicines longer term. We're also doing studies in the younger patients, and so we're able to treat these children at much younger ages of their lives. And so our commitment and investment to CF and these medicines we believe will be hugely beneficial to that is maintained. So you could think about that as a kind of a steady-state part of our R&D investment. And then as we think about advancing our pipeline in other disease areas beyond cystic fibrosis, we are happy to be letting you -- we'll bring you more up-to-date with how we're advancing new medicines into the clinic in these new disease areas that Jeff just referred to earlier. The investments in these disease areas is smaller in magnitude. It's earlier stage, but it is additional investments. And therefore, when you compare a 2018 operating expense profile to a 2019 operating expense profile, you should see a relatively small increase compared to the increase we anticipate on the revenue line. So yes, we do anticipate operating to increase. And then one last thing I would say is that, you have to remember, we're also launching medicines globally, not just where it's been in the U.S. And therefore, there are launch costs that are being incurred outside of the U.S., not significant because of the nature of the disease and how we can reach these patients with our medicines, but there is investments there as well. So you should expect to see operating expense margin increase in 2019 compared to 2018, but the key will be the revenue growth, which will allow us to continue to generate more cash flow.

Our next question is from Michael Yee from Jefferies.

On the once daily, you made a comment that you've had some positive feedback and are moving that forward. Can you just, I guess, clarify the process going forward and why you're choosing gating mutations as sort of the first move forward and not maybe a broader population. How should we think about that? And then the second question was more of a modeling finance question, which is that, obviously, the triples are advancing forward like you said, but ORKAMBI and SYMDEKO are pretty big franchises. Over some period of time, is that expected conceptually to have those doublet drugs go away and the triple is basically, is the single line item? How do you think about that? Are there patients that actually would stay on doubles?

This is Reshma. Let me start by taking the first part of the question with regard to VX-561. As you know, and we've talked about this on previous calls. The FDA had asked us to do additional dose ranging before proceeding to Phase III. And part of that request was to study VX-561 across these dose ranges in gating patients. And we've had productive discussions with the agency and we're now at the point where we've designed our study, we selected our doses, and I expect that the study will get started next year probably in the first half or so. Stuart?

Yes, Michael, on the question of what do we think the triple combinations might do to the doublets. Certainly, if we see the sort of results in Phase III that we saw in Phase II in the homozygous population where, as you know, we saw the triple combinations deliver very, very substantial increases in clinical benefit over and above a baseline of tezacaftor and ivacaftor. If those results hold true, then our expectation would be that the vast majority of patients would be transferred to a triple combination rather than staying on a doublet just given that the significant benefit-risk improvement that we saw in those Phase II studies.

Okay. And just to clarify on 561, you have to test it in gating patients first, but the full plan certainly is to move it to a broader population after that. Is that fair?

You got that right.

Our next question is from Robyn Karnauskas from Citi.

So I guess let me ask about VX-150. So you mentioned that you're doing a dose-ranging study. You've already done some trials in acute. I was curious if this is to get a drug that may be superior to Vicodin since the other doses were shown to be equivalent. And then another question on that is that you talked about the acute market. And some checks we've done suggest that reimbursement might be challenging there because doctors may just give a cheap generic drug. What has your market research told you what has to happen for a more expensive, nonaddictive drug will have uptake in that setting?

Yes. Thanks for the questions, Robyn, both good questions. Let me start with maybe your second one. I'll come back to the dose-ranging study which is pretty straightforward. With respect to reimbursement, you're correct, a large part of the acute market has been genericized. And as you know, a large part of that market is actually opioids for acute postsurgical pain, post-dental pain, et cetera. Obviously, we're sitting in the middle of an opioid epidemic and the #1 objective that, I think, patients, physicians and the government has is to find medicines that have essentially equivalent efficacy to opioids without the abuse potential and the side effects of opioids. We believe that VX-150 based on the data we have so far has the potential to hit that profile. You saw that in the comparator trial that we did. If that's true, we feel that VX-150 is a very, very competitive drug to a generic opioid. And with reasonable pricing, that market is a multibillion-dollar opportunity even if you only capture a portion of it. In the U.S. last year, there were over 200 million opioid prescriptions written, which is a truly astounding number. Obviously, not all of them for acute pain, but a lot of them. So does that answer the question in terms of reimbursement?

Yes, that's very helpful.

Okay. Then in terms of VX-150. I'm sorry, go ahead. Did you have another question?

No, go ahead. I'm just asking about the dose ranging, if your dose is higher if you maybe show a superior profile in the acute setting, this may make reimbursement even easier.

Yes, yes. So you may remember the study that we did in acute bunionectomy is essentially, I would call it a proof-of-mechanism study because it was the first time anybody had shown that a 1.8 inhibitor could work in acute pain. And so we actually used a fairly high dose there just to make sure that we were getting up to a reasonable exposure. We also know that before you can move into Phase III, regulators will require you to do a dose-ranging study, really to understand 2 things. One, what is the PK/PD profile of the drug. And frankly, two, what is the lowest effective dose because you always want to give the lowest fully effective dose. And so think of this as just the standard part of the Phase II regulatory process where we need to identify the PK/PD profile or the lowest effective dose. And that would then enable us to go to the agency with a plan for a pivotal trial, again, assuming that it's successful.

Got it. And second question on alpha antitrypsin. So you mentioned the sweat chloride, there is more sweat chloride that you see per mil, so efficacy. But I do have flashbacks from sweat chloride with ORKAMBI, I think, not correlating with FEV1. In this indication, is it more straightforward that if you see a reduction, you're going to see an improvement in efficacy? Help us understand a little bit the disease.

Sure. Sure. Yes, let me maybe give you a little context on the disease, if you don't mind my taking a minute. One of the remarkable things about AAT is how similar. It sort of smells and looks a lot like CF in many, many different ways. It's a genetic disease. It affects somewhere between 60,000 and 100,000 people in the U.S. and probably about another 50,000 to 60,000 outside the U.S. It's caused by one major mutation in more than 90% of people. And it's a protein folding disorder as it turns out, just like CF is with sort of remarkable similarity. In this case, the normal protein AAT is made in the liver and is normally secreted into the blood where it travels to the lung. And it protects the lung against autodigestion essentially by endogenous enzymes. So in the disease, the mutant protein has 2 problems. One, it doesn't fold correctly, it accumulates in the liver, and so it causes significant liver disease actually in up to 30% of all the patients. And two, obviously, it's not secreted into the blood so it's unable to go to the lung and protect the lung. And so the lung gets autodigested and these patients suffer from early-onset emphysema. So that's the disease. The challenge is for us was to, could we make a corrector, a small molecule corrector of AAT, just as we have for CFTR, that would essentially refold the protein in the liver, so be secreted from the liver and therefore presumably help treat the liver disease. And more importantly, that would secrete functionally into the blood where it would go to the lung and protect the lung. That was the challenge. It's a very high problem, I can tell you biological problem. I think we've cracked that problem now. We've made a series of small molecule correctors that clearly do both of those things. They refold the protein so it's secreted, and it's secreted in a functional form so that it could protect the lung.

One of the really nice things of this disease is there is both a cell or several cell models and there's also in this case an animal model, which we didn't have in CF in which the human mutation is then engineered into the animal. And so we've been able to test these molecules both in the cells and in the animals, and they do what we thought they would do. They refold the protein, secrete it from the cells. In the animals, they secreted AAT at levels that would be sufficient to essentially be preventative or curative in the disease.

And then the other thing that's very similar to CF is that it's a nice biomarker, which is simply the circulating levels of AAT in the serum. You can treat the patients for several weeks and just measure both levels and function of the AAT in the serum. And we know from human -- experiments of nature essentially that if you have a certain level of AAT in the serum, you don't get the disease. And so we know exactly what we're shooting for here. And we can measure it very easily. And so the plan is to take this into normal healthies by the end of this year, the first of these molecules, and we have a whole series of them, just like we did with CFTR correctors. And then if they are safe and more tolerable, once we understand the PK, take them into patients. And again, with relatively small numbers of patients treated for relatively short periods of time, you simply measure AAT in the serum and you know exactly where you are.

Our next question is from Geoffrey Porges from Leerink.

First, Jeff, you mentioned a mid-teens FEV1 improvement now being the hurdle for triples for competition. Could you give us your sense of what the bar is for sweat chloride relative and absolute for homozygous and het/min so we can know how to appraise all these competitors coming along? And then just wondering, Stuart, if you could give us a sense of where you are in the eligible Delta F508 penetration in the U.S. between SYMDEKO and ORKAMBI. You've sort of mentioned that you had quite a few discontinuations in patients who never started on ORKAMBI. And have you called all those patients up yet or are you still substantially below the penetration you've achieved with KALYDECO in the eligible paying population spectrum?

Yes, Geoff, thanks, I'll answer the first part and Stuart then, of course, will answer the second part. And first of all, I do want to just correct what may have been a misimpression of what I said or maybe I said the wrong thing, but I want to be very clear. I don't think mid-teens is the bar, for instance, for het/min patients. I think the bar that we've set, as you know, from the data is somewhere between 9.8% and 13.8% in our Phase II data for het/mins. And for homozygous patients who already have an effect from doubles that we estimate is around 4%. We show an additional 9.6% or 10% improvement, which is how I get to that mid-teens. So just want to be clear that I'm not suggesting that 15% is the bar that you need for het/mins to get to be successful. With respect to sweat chloride, I think these are approximate numbers, but I think when you reach sweat chloride that approaches carrier levels, which is our sort of goal, let's say, 40 to 50, that's probably the goal that you know you're there with the triple. And so I think those are the 2 numbers we look at. In most cases, but not in all cases, they correlate pretty well. And at the end of the day, as you know, the regulatory endpoint is going to be, I think, the one.

Geoff, it's Stuart here. Just commenting on the SYMDEKO launch in the U.S. So as you suggested, one of the most important things for us for SYMDEKO is the ability to get patients who were not being treated by ORKAMBI onto a CFTR modulator, and that was really 2 populations. One was those who were completely naïve, have never been exposed to ORKAMBI. And we've seen good uptake in that population. And the second population was those who had discontinued ORKAMBI often because of the respiratory adverse events. And we've also seen good uptake in that patient population as well. We've also seen a number of patients switch from ORKAMBI to SYMDEKO. So we've seen strong uptake in all those patient populations. We're certainly not done with the launch. Whilst we've seen good uptake, Geoff, we're certainly not done. We haven't flattened out on SYMDEKO uptake in the U.S. yet. So we still have some ways to go, but we're certainly pleased with how the launch has gone so far.

And Geoff, maybe just to add one thing that I know you're aware of is. You talked about the efficacy side of these triples, but clearly tolerability and safety is the other side of the equation, and long-term tolerability and safety, not 1 week or 2 weeks. And so I think that's going to be an important thing that we all look at as well as these trials progress.

Our next question is from Terence Flynn from Goldman Sachs.

I was just wondering first, you might not comment on this yet at this point, but just with respect to the triple combo. Can you maybe just give us an update on the different inputs you're thinking about with respect to pricing? And then I would just love your latest thoughts on capital allocation here. I know Jeff you made some comments in your opening remarks. But you have $3 billion on the balance sheet that might be the most you guys have ever had. So just curious how you're thinking about that as that continues to grow here into next year.

Yes, Terence, Stuart here. On pricing for the triple, obviously, much too early to make any specific comment, but we'll be taking into account the same factors that we always do. That will be the magnitude of the clinical benefit that we're able to deliver for patients. It will be the label that we get approved for and that obviously leads into what's the eligible patient population. So those are really going to be the inputs that we take into consideration. And obviously, we'll get more information on those in the coming months.

And as far as capital allocation is concerned, it's really interesting. I can tell you that being in the company has invested mainly and have been about capital preservation for a long period of time and then in the last 3 years, we're now a cash accumulator and creator. And it's about capital allocation. It is an interesting philosophical change for the company, but I'd like to start by saying the revenue line is the main source of cash because after that, we then allocate inside the company on our internal programs. It's often not told about because of capital allocation, but the main source of capital is our revenue line and then we allocate that internally. And as you can see in terms of the questions asked on operating expense earlier, we still allocate a good amount of capital towards cystic fibrosis medicines, but also now we're starting to allocate capital to diseases that are beyond cystic fibrosis, and we're really happy about that. Some of those opportunities that are beyond cystic fibrosis are our next order of priority, which is we're allocating capital outside the company towards collaborations and towards medicines that we can accelerate into development. And some of the opportunities we have with beta-thalassemia and also sickle cell have been a product of a very successful and nice collaboration with CRISPR Therapeutics. And we're really excited to be going into the clinic very shortly with these medicines for sickle cell and beta-thal. And we anticipate that there will be others to follow. The medicine for FSGS, for example, was something that we in-licensed from a very early stage a couple of years ago. And what it's allows us to do is to really reignite a research just in that area is now providing the opportunity to go into the clinic in a very short period of time for a very important medicine. So it's a high order priority for us to allocate capital outside the company to do collaborations to give us opportunity with more modalities and more medicines for diseases that we would be interested in. And I would say, there is also a consideration for how we may also limit the increase of our share count. Our outstanding shares does increase year by year based on the issuance and exercise of options and restricted stock. And so we like to think of ways that we may limit that creep in outstanding shares. And we've been doing that throughout this year. We did announce earlier this year that we entered in a $500 million share buyback program. And we have been buying shares back throughout this year. And that's been very successful for us as well. And therefore has limited the creep in the outstanding shares, which obviously helps us with the small amount on the dilution. And so that's how we think about capital allocation. It's very important, very close to us. And we're happy that we have the opportunity to prioritize it in this way.

Our next question is from Alethia Young from Cantor Fitzgerald.

This is Eliana for Alethia. Just on the earlier side of things on the R&D side, you recently did a deal with Genomics. So could you talk a little bit more about this and sort of how having this particular discovery engine will sort of enhance your R&D efforts.

Yes. Thanks for the question. We were actually quite excited about that collaboration. As you know, one of the principles of our research effort is to really work on what we call highly validated targets. There's really 2 major ways we validate targets. One is pharmacologically with medicines. That's actually the minority by the way. And the second is human genetics, where mutations, either gain of function or loss of function would tell you a lot about the role of specific targets. A beautiful example of that is NaV1.7 and NaV1.8, which we validated by human genetics, obviously, CFTR, et cetera. The collaboration with Genomics, who we believe is the leading small company in the world working on this kind of high throughput human genetics, will allow us to screen for new targets on the one hand through large scale data analysis of human genetic databases. But also it will allow us to ask questions for instance about the potential side effects of hitting a specific target because you can screen the human population for mutations on that target, a loss of function, for example, and ask other people normal or not, do they suffer from any mechanisms or side effects. That's a very important thing to know as we begin to pick targets and develop small molecules. So think about it as a way of tapping into the human genetic data resource that both help pick targets and help screen for safety. And that collaboration is off to a very nice start in a number of different diseases.

Got you. And then just another one in terms of early-stage development. Can you just tell us a little bit more about the preclinical models for AAT? I mean, obviously with CF, you guys are very successful in optimizing these preclinical models. But could you tell us a little bit sort of about your thoughts on what you think the best preclinical model for AAT is and also sort of your thoughts on how predictive these models might be?

Yes. Thanks for the question. There's 2 kinds of models that are generally used. These are not, by the way, necessarily proprietary to us. One is a set of cell-based models in which either directs from patients are derived from engineered cells with the human Z mutation in them, which is the major AAT mutation. And in those cells, the mutant protein is made, it's misfolded, it's not secreted. And so one can test molecules on those cells as to the help refold the protein, secrete it as a functional, et cetera. The other one is that there's a transgenic mouse which actually expresses the human mutant protein. Again, that protein is not secreted well. It accumulates in the liver of those mice. And once we dose those mice with medicine, you'll have these protein secreted, it's a function of what levels do you reach, et cetera. So 2 very nice models which we feel actually will be highly predictive because they do actually study the human mutation, not by mouse-driven mutation.

Our next question is from Ying Huang from Bank of America Merrill Lynch.

First one, maybe on again competition. We have seen some data from Proteostasis. I was wondering if you guys have done any work in the lab about the role of amplifier. Do you have a view on that or not? And then secondly maybe for Stu. Can you tell us by when you might see SYMDEKO pretty much reaching a full penetration in the U.S. market given where you're seeing the market today?

Yes, this is Jeff. I'll take the first one in terms of the amplifier. Again, we really don't comment on other people's specific molecules. One of the things that we like about our triple and our correctors are, we know that they're quite specific for the CFTR protein, which is I think one of the reasons, quite selective, if you will. They don't hit other proteins. One of the reasons why they've been so tolerated. And so I think as we look at -- as you look at other molecules, that issue of specificity imposed is an important thing to look at.

And then in terms of the SYMDEKO uptake. Obviously, we're very pleased with how it's gone to-date. Predicting exactly where it's going to reach full penetration is hard to do, but you can rest assured that when we get to the point of giving '19 guidance, we'll be including our expectations in that for how SYMDEKO is going to perform in the U.S. and then also hopefully by then, we'll have the approval of SYMKEVI in Europe as well and we'll incorporate all that into our '19 guidance.

Our next question is from Brian Abrahams from RBC Capital Markets.

In the NACF data, we saw more specifics on the rate of cough with the triple combo. I know these weren't classified as bronchoconstriction. I was wondering if you can give us any more details in the quality, severity and duration. Remind us, should we think about this as related to the initial mucus clearance and how it might compare to ORKAMBI? And then just a follow-up on the pain program. Obviously, some really interesting Phase II data. Wondering if there are ways you might be able to further improve upon the bioavailability and potency of VX-150. And should we assume 128 is no longer being pursued?

Yes. Maybe I'll take both of those real quickly. I think the question you're trying to get at for the triple, but correct me if I'm wrong is, are we seeing bronchoconstriction as we saw with ORKAMBI? And the answer is no. And we've actually measured that specifically not only in terms of cough or anything else. So we do not see bronchoconstriction with the triple, and we did not see it with tez/iva as you remember. But I think the more important point, right, Brian, is just looking at the total tolerability and safety profile of these drugs now in a couple of hundred patients. What we're seeing is they're quite clean. They look a lot like the earlier drug, ORKAMBI accepted. They don't have the bronchoconstriction means that they've been very well tolerated and very safe so far. Did that answer your question? I just want to make sure that was the question you're asking.

That's very helpful. Yes. Just wanted to contextualize that, yes.

Yes. And then the other question was about 150 and 128?

Yes.

Yes. So maybe just again, let me take a step back and remind you that the approach that we're taking in pain, actually not unlike the approach we've taken in CF and the approach we're taking in AAT, is to create a portfolio of molecules, not one, not single rifle shots, but multiple molecules that have different properties, both with respect to things like potency, but also with respect to things like PK, biodistribution, drug-drug interactions, et cetera. And then take those molecules forward into early-stage clinical development to understand their profiles more fully. Obviously, 150, we have the most data on. We know that it's well tolerated. We know that it has efficacy, at least at these higher doses in acute pain and in osteoarthritis. And we'll get a readout on the third kind of pain, neuropathic pain, early next year. 128 was a follow-on molecule, if you will, with some different properties than 150. We discontinued the development of 128 in Phase I in healthy volunteers because it didn't have the PK profile, the tolerability improvement that we were hoping to see over 150. We have multiple additional 1.8 inhibitors coming behind that. And so you should expect to see us bring other inhibitors into early clinical development, the same way trying to optimize these molecules. Remember, in acute pain, you're going to need molecules that are highly efficacious and also highly safe and highly tolerable. We don't want to create molecules that have liabilities like opioids.

Our next question is from Paul Matteis from Stifel.

This is Ben Burnett on for Paul. I wanted to ask another question about VX-150. I wonder if you can provide a little bit more color on the Phase IIb study in acute pain, like specifically what are you looking for and what is the efficacy hurdle that you need to advance this to a pivotal program. And then the second question just sort of more at a high level. I guess across these different pain indications, the acute and the neuropathy is ongoing as well as osteoarthritis-related pain, you have some data there. What are some of the ultimate vision for this program? And I guess under what scenarios would you want to commercialize this yourselves? And is there a scenario where a partner might be involved in that aspect of it?

Yes, Ben, thank you. I'll take the second part actually first, then gets a little more strategic, maybe we'll give you some context and then Reshma will take the first part, which is about the VX-150 Phase IIb trial. I said this before, but maybe just to give you some context. We think about pain both from a commercial standpoint and from a scientific standpoint not as one disease, but is actually multiple diseases. So an easy way to divide this, for instance, is acute pain. That's the kind of postsurgical pain, the dental pain when you have an acute injury. That's one segment. The second segment is chronic inflammatory pain, things like osteoarthritis, lower back pain. A third segment is neuropathic pain. That's pain like diabetic neuropathy. They have different commercial channels obviously, but they also potentially have different pain mechanisms. And so it's important to study them separately because some drugs will work in one and some drugs will work in 2, and frankly some drugs might work in all 3. So that's the approach that we've taken with our portfolio. We've done separate trials in acute pain using bunionectomy as the model. We've done separate trials in chronic inflammatory pain using OA as the model, osteoarthritis. And we're doing a third set of trials in neuropathic pain using small fiber neuropathy as the model. And as you know, what we've seen so far just to remind you is, a high level of efficacy in acute pain, a high level of efficacy in osteoarthritis chronic pain, and we're still waiting to see the data from the neuropathic trial, which is due early next year. So that's where things sort of stand from a scientific standpoint. Now again if I take a step back, remember that there is a very large need for a new class of pain medicines. There really hasn't been a new class for over 50 years. And the reason is that we have opioids on the one hand which have high efficacy, but are burdened with all of the addictive and other side effects. And then we have drugs like aspirin, Tylenol, NSAIDs, which have lower efficacy and also by the way, have some of their own side effects which are not trivial. And so the target for us is really can we find a new mechanism of pain that approaches opioid-like efficacy without the liabilities, addictive and other liabilities of opioids. And so far, there's been Phase II data, what we're seeing is VX 1.8 looks like such a target. Sorry, NaV1.8 looks like such a target. VX-150 is the first medicine that does look like it hits in both of those 2 pain indications. And the third one we're waiting to see. And the final part of your question is, what about commercializing that. It turns out, acute pain, as I said, is a multibillion-dollar opportunity and at least many, not all necessarily, but many of the channels in the acute pain market can be addressed by a specialty sales force. That's something that we would develop and commercialize ourselves. Contrast that with, say, OA pain or low back pain, that's clearly not a specialty disease. That's a community disease or requires a primary care sales force. And while we may take the science and Phase II trials forward, we will certainly not develop our own primary care sales force and commercialize that. That would be done with a partner. Nevertheless, we think, if we have such a molecule, it's a very valuable asset for us to partner. And the neuropathic pain lies somewhere in between. It can be covered by a rather large specialty sales force, meaning 100 to 150 reps probably in this country. So it is mostly by specialist. And it really depends on the quality of the data that we're going to see there in the first part of the year in terms of deciding what we would do with respect to further development and commercialization. Does that give you sort of a lay of the land of how we're thinking about it?

Makes sense. Appreciate it.

Okay. Reshma?

This is Reshma. With regard to the VX-150 Phase II study that we are going to be initiating shortly, it's a fairly standard Phase II dose-ranging study. This is going to be the bunionectomy acute pain model. And in essence, it's going to be several doses that we study to get a full dose exposure range. You'll remember that the bunionectomy study that had the positive results from early in the year was a high-dose study because we were really testing proof of mechanism in that one. The endpoint is going to be SPID24, SPID48. This is a pretty standard endpoint for the bunionectomy model. And with regard to the treatment effect, we were very pleased with the treatment effect we saw in the bunionectomy study. And now we need to explore the full dose range and exposure to see what the lowest effective dose is going to be. And that's really what we're studying here.

Our next question is from Cory Kasimov from JPMorgan.

Two for me as well. In your prepared remarks, you mentioned you'll have some enhanced next-gen correctors moving into the clinic in the coming months. I'm wondering if there's any additional information you can provide about these in terms of how they might be further differentiated from what you already have. And then my second question is with regard to the clinical hold for CTX001 in sickle cell now being lifted and initial trials slated to begin. How much data do you ideally want to accumulate before you're potentially comfortable providing an update? And will this be something that's your responsibility or your partners?

Yes, Cory, this is Jeff. Maybe I'll take Part 1 and Ian will talk about disclosure and Part 2 of the question. With respect to the enhanced next-gen correctors, I would say what we're looking for here is 2 things. One, I think we've shown you the chart before that as we've gone through the last 4 years, it's really remarkable to see how we've been able to improve the potency and efficacy of these next-gen correctors. With the current molecules, 659 and 445, we're close to getting all patients to carrier levels, but not quite there. And so the goal would be to find a next-gen corrector that could get all patients, meaning het/min in particular as well as homozygous patients to carrier levels of chloride transport because we believe if you can get there, particularly in young patients, you'll either prevent or greatly, greatly lessen the severity of the disease. The other goal of that program is to continue to diversify the chemotypes. We always like to have multiple chemotypes of any medicine that we're making. And so we've been able to explore alternative chemotypes, and we're working on those as well. So those are the 2 kinds of molecules that you can expect to see entering the clinic. We already have examples of both. And so the first one to go, I do expect one in the clinic early next year. And you can probably see some more as the year goes along as well. Does that answer your question about the enhanced next-gen correctors?

Yes. Perfect.

Okay. And Cory, as to your disclosure question. Well, we just started recruiting patients for these studies. And I think the next disclosure we'll provide you with would be kind of an update of where we are on the recruitment. We'd anticipate dosing first patient probably towards the end of this year, early next year. And then once we have a number of patients in the study, we'd anticipate a disclosure of data from the study later on next year, but that would probably be timed with multiple patients in both studies and being on therapy or let's say, have had the procedure for a certain duration of time.

Well, that will conclude the call. The Investor Relations team is in the office tonight to answer any further follow-up questions. Thank you very much for joining the call and have a good evening.

Ladies and gentlemen, thank you for your participation in today's conference. This concludes the program. You may now disconnect.