FibroGen, Inc.

Thank you all for joining on the call. With [indiscernible] Executive Officer of FibroGen, Dr.

Peony Yu, Vice President of Clinical Development; Dr. Frank Valone, Chief Medical Officer; Pat Cotroneo, Chief Financial Officer; and Dr.

Lynda Szczech, Executive Director, Clinical Development.

On this call, we expect to make forward-looking statements regarding our business, including our collaboration with AstraZeneca and Astellas, financial guidance, the initiation enrollment, design conduct and results of clinical trials, research and development activities and certain other statements regarding the future of our business.

Because forward-looking statements relate to the future, they are subject to inherent uncertainties, risks and changes that are difficult to predict, many of which are outside of our control. We refer you to the Risk Factors section of our annual report on Form 10-K, for the year ended December 31, 2014, our quarterly report on Form 10-Q for the quarter ended March 31, 2015, and our quarterly report on Form 10-Q for the quarter ended June 30, 2015, each of which has been filed with the SEC.

For risks and uncertainties regarding our business as well as the statements made on the call today.

We undertake no obligation to update any forward-looking statement, whether as a result of new information, future developments or otherwise. A webcast of this conference call will be available for replay on the Investor Relations page at FibroGen's website, www.fibrogen.com.

I'll now hand the call over to Tom Neff.

Thank you, Greg. Good afternoon.

On today's call, we will review key updates, discuss accomplishments and results during Q2, and highlight our most important near-term and long-term goals. Many of these on the call today may be new to FibroGen, so I'd like to begin with a brief introduction.

The company is advancing 4 distinct clinical stage programs in anemia, fibrosis, cancer and corneal implants. In anemia, we have our lead compound, roxadustat, a small molecule investigational therapy based on hypoxia inducible factor system, which has been shown to trigger the natural responses to anemia.

Roxadustat is currently in Phase III trials in the U.S. and in Europe, and is now expected to interface 3 in China Q4 this year.

Roxadustat was discovered and developed by FibroGen and is partnered with AstraZeneca and Astellas under multiple exclusive territory-based agreements. Our roxadustat patent portfolio includes multiple issued and pending U.S.

patents, offering protection for roxadustat through 2033 and potentially beyond.

As noted in today's press release, we remain on track to achieve our goal of submitting regulatory applications in China in 2016 and in the U.S. in 2018.

Working closely with AstraZeneca and Astellas, we have launched all 7 studies required for the roxadustat regulatory submissions in the U.S. and EU.

4 of these studies are in hemodialysis patients where total enrollment is planned for just over 3,500 patients, 3 of these studies are in non-dialysis with target enrollment of approximately 3,800 patients. Of these 7 studies, 3 are being run by FibroGen, 2 by Astellas and 2 by AstraZeneca.

The 3 FibroGen studies are focused in the U.S. and account for 25% of patient enrollment.

The 2 Astellas studies are focused in Europe, the 2 AstraZeneca studies are global in character and much larger than the others, accounting for more than 1/2 of the entire program, with target enrollment of more than 4,000 of the 7,300 patients we need to complete Phase III enrollment.

For the 3 FibroGen roxadustat Phase III studies, we have now reached approximately 2/3 of our cumulative target enrollment. During the past 3 months, we have been rolling in the range of 125 to 150 patients per month in our 3 studies.

The enrollment run rate give us added confidence that FibroGen can meet our enrollment goals.

As was agreed with our partner, AstraZeneca, at the beginning of this year, the base plan is for FibroGen to complete target enrollment by March to April 2016. The stretch goal for all of our studies and their studies is to achieve full enrollment by year-end 2015.

At present, we expect one of our studies to meet or beat the stretch goal. And for the other 2, we now think we can meet or beat the March, April targets.

As we previously stated in our March call, we completed 2 separate 2-year carcinogenicity studies, for roxadustat, 1 in rats and 1 in mice. These studies showed no evidence of effect on carcinogenicity or mortality.

And as we announced on May 7, the audited study reports triggered a $15 million payment from AstraZeneca for this milestone achievement.

In the roxadustat clinical trials, we have seen aspects of a clinical profile that are unique compared to previously approved anemia therapies. Roxadustat increases hemoglobin levels within or near physiologic level, using within or near physiologic level of endogenous epo regulates genes that make iron bioavailable.

And in addition, evidence shows that roxadustat suppresses the iron regulatory hormone, hepcidin upregulates the iron transport factor transferrin. Hepcidin is upregulated into flammatory conditions and blocks access to iron and macrophages that is used to make red blood cells.

Studies have shown that IV iron increases hepcidin levels, which may interfere with the activity of the approved anemia therapies. Transferrin is the main protein that binds to iron and transports it through the body.

Consistent with these unique actions in our Phase II studies, we saw that roxadustat's ability to correct anemia does not appear to be impaired by inflammation irrespective of the level of inflammation. The paradigm is thus very different than recombinant epo and IV iron, where the presence of inflammation requires increasingly high doses.

The data suggests that roxadustat may be able to overcome inflammation to treat dialysis patients who are hyporesponsive to currently approved therapies, with erythropoiesis stimulating agents and IV iron as well as potentially to treat patients with inflammatory bowel disease, lupus nephritis, ulcerative colitis and other such categories where ESA products have not received a label.

We are continuing to evaluate the possibility of developing roxadustat for these and other additional indications. Based on these data and our differentiated mechanism of action and our discussions with our partner, AstraZeneca, including the view that the inflammatory anemias are considered unmet medical need.

We modified our company disclosure during the second quarter of 2015 to indicate our assessment that for U.S. dialysis patients, it is now more likely than not that roxadustat will be separately reimbursed, rather than being in the dialysis drug capitation bundle as was contemplated at the time of our IPO.

The global scope of the roxadustat program includes an agreement with our partner, AstraZeneca, for the development -- commercialization of roxadustat in China. We have been informed that the required technical review of our Phase III clinical trial application by the Center for Drug Evaluation, otherwise known as the CDE, has been completed and the file has been referred back to the CFDA to obtain the official authorization to conduct the Phase III program.

We are in -- thus we are now at a stage in our China program where we are advised that the remaining regulatory approval steps to start the Phase III studies are procedural. Therefore, we now expect to begin opening study sites and enrolling patients in 2 Phase III studies in China in Q4 2015.

As those of you who were involved in our IPO may recall, we committed to publish 6 manuscripts period-to-period journals in the first half of 2015, all have now been submitted and the publications of the roxadustat Phase II studies are beginning to roll out. As we announced yesterday, the placebo-controlled dose rangefinding study in non-dialysis CKD patients, Study 017, has just been published in the current online edition of Nephrology Dialysis and Transplantation, or NDT, the Journal of the ERA-EDTA in Europe.

We expect a second manuscript will be published by a major journal in the coming weeks, along with editorial comments and the other studies will be impressed during the months ahead.

These are the first peer-reviewed clinical studies to feature clinical evidence of hemoglobin correction by a hypoxia-inducible factor prolyl-hydroxylase inhibitor in anemic patients, the technology where FibroGen has been the innovator of record. We are very proud of this milestone.

But, at the same time, it is taken a team effort. We want to recognize the tireless support of the teams at AstraZeneca and Astellas who have reviewed CSRs, the study data and the manuscripts, and extend our congratulations to the authors, particularly to our academic collaborators who have made critical contributions over the past decade in many ways.

And fibrosis, our second major clinical program, CTGF was codiscovered by a former FibroGen scientist in NIH research team. And it has since been shown that time after time, that CTGF is a central mediator of fibrosis.

As a result, we hold proprietary patent positions around the blockade of CTGF to treat fibrosis. We focus on development of FG-3019, a fully human antibody that was generated using Medarex technology and that blocks the biologic activity of CTGF.

There are particular preclinical results that drive our plans in exploring the use of FG-3019 to treat human fibrotic diseases. In proof-of-concept experiments we showed that FG-3019 can't reverse radiation-induced fibrosis.

Mice received a lethal dose of radiation to their chest, after which the mice developed progressive lung fibrosis that, without effective treatment resulted, most mice dying within 1 year.

16 weeks after the lethal radiation dose, the mice had substantial fibrosis and at that time, we began injecting FG-3019 treatment an antibody 3x a week, for a total of 8 weeks.

Lung fibrosis improved significantly within 2 weeks of starting FG-3019 and this reversal persisted even when treatment stopped after 8 weeks. There was a significant mortality difference between the treated and untreated animals.

Subsequently, we tested the ability of FG-3019 to reverse fibrosis in idiopathic pulmonary fibrosis patients. We monitored lung fibrosis using high-resolution computed tomography, or HRCT, and quantified lung fibrosis using a computer algorithm developed at UCLA.

The results of our study showed that 35% of subjects had stable or improved lung fibrosis and lung function after treatment with 3019 for 48 weeks. To our knowledge, FibroGen is the only company to report improved lung fibrosis in an IPF clinical trial.

We are now conducting a placebo-controlled trial in IPF patients to evaluate further the safety and efficacy of FG-3019. As part of the previously announced expansion of this study, we began opening new trial sites outside the U.S.

in the third quarter of 2015 to help accelerate enrollment. As Dr.

Valone will explain, we are considering now to expand this trial to include a cohort of subjects who are receiving approved therapies in the United States. We believe that the results of this trial if positive will point the way to further product development Phase III trials.

Another area of fibrosis, we are very excited about our Duchenne muscular dystrophy program. We have been working on this program for several years to validate the CTGF mechanism of action in the disease and to test FG-3019 in DMD preclinical models.

More than 20 papers based on this work have been published in peer-reviewed journals by our collaborating investigators. As a key step towards moving into the clinic, we met with the International Muscular Dystrophy Review Board, TREAT-NMD, Advisory Committee for Therapeutics and reviewed these data and the key results.

The TREAT-NMD review committee was supportive of conducting studies with FG-3019 in DMD patients. Since then, we met with American and European KOLs to review these findings, discuss the rationale for FG-3019 as a potential treatment for DMD as well as other muscular dystrophies and to refine the clinical trial design.

We received FDA clearance for our IND, for DMD last month and plan to initiate a Phase II study of FG-3019 in non-ambulatory DMD patients in Q4 2015. We also plan to meet with FDA to discuss requirements for treating ambulatory patients, particularly those aged 3 to 11, shortly.

The third FibroGen platform is in cancer. Here, we are exploring the impact of altering or reducing fibrotic extracellular matrix on tumor growth metastasis.

The results from our first dose escalation study conducted in pancreatic cancer patients, showed that achieving higher circulating blood levels of FG-3019 was associated with an improvement in survival in pancreatic cancers. The efforts to explore the meaning of this result in pancreatic patients continue both on a clinical and on a molecular level.

We are now conducting a Phase II study in Stage III pancreatic patients who have been scored as not eligible for surgical resection.

The study examines whether the combination of FG-3019 and the standard chemotherapy regimen of gemcitabine and ABRAXANE plus radiation can convert these subjects' tumors from inoperable to operable state. Patients with inoperable Stage III pancreatic cancer have survived similar to patients with metastatic cancer with studies showing only half of the inoperable Stage III patients being alive approximately 8 to 12 months after diagnosis, and very few studies even report 5-year survival.

The outlook for patients with resectable pancreatic cancer is considerably better, with studies showing that half being alive between 17 and 27 months after diagnosis, and 20% alive at 5 years. We intend to enroll up to 40 patients to determine if there is a significant difference versus standard of care alone and then make a decision to expand the study.

Dr. Frank Valone will provide further updates on IPF, DMD and pancreatic programs later in this call.

Our fourth platform is the use of recombinant human collagen for corneal implants as a potential treatment for corneal blindness.

Capitalizing on our expertise in the biology of fibrosis, FibroGen developed the unique capability to make human collagens of many different types. Early on, in that program, we realized we could fabricate thermally stable Type III collagen into a corneal implant that forms a clear crystalline structure that had optical properties similar to the human cornea.

In the human pilot study conducted in Sweden, and now in its seventh year, corneal implants made of our material were shown to restore sights successfully in patients with corneal blindness. In this pilot study, the human collagen III implant integrated well with the native tissue, with nerve regrowth observed, patient sense of touch restored and no tissue rejection.

Importantly, these data indicate that this therapy should not require extended immunosuppressive supportive care. We are proceeding with the planned development activities of a 5200 molecule as a Class III medical device in China for partial thickness and full thickness procedures, where there are limited treatment options for corneal implants.

There are an estimated 4 million to 5 million patients in China with corneal blindness. Finally, I wish to mention a few items related to our financial position.

As anticipated in the second quarter, we received a milestone payment and non-contingent license payment from AstraZeneca, totaling $135 million. Under the 50-50 cost share agreement with AstraZeneca, based upon agreed budgets and current enrollment forecast, we expect to reach our cap of $116.5 million by mid-quarter 4 2015.

After the cap is reached, AstraZeneca and Astellas reimbursed our agreed-upon cost related to CKD anemia, with the exception of China, where we continue to share 50-50 with AstraZeneca. At June 30, 2015, FibroGen had approximately $408 million of cash, cash equivalents, investments and receivables.

Now we will turn this call over to our lead clinicians, Dr. Peony Yu, in the anemia program and Dr.

Frank Valone in fibrosis and cancer for some details on the trials. Then our CFO, Pat Cotroneo, will go over the numbers and we will conclude the session with some questions.

Peony, please go ahead.

Thank you, Tom. I'd like to comment briefly on the profile of roxadustat, and its potential for treating patients with anemia, and then describe our Phase III program.

Roxadustat is our first-in-class hypoxia-inducible factor prolyl-hydroxylase inhibitor in development for the treatment of anemia in patients with chronic kidney disease. In response to hypoxic conditions, such as high altitude of blood loss, our body naturally makes more red blood cells to carry oxygen.

Roxadustat pharmacologically turns on that response system to make red blood cells.

Tom already mentioned how roxadustat's mechanism of action includes mobilizing iron availability, reducing hepcidin and stimulating only a modest amount of the body's own erythropoietin. We believe a novel approach with roxadustat leads to a differentiated therapeutic profile and potential advantages over the class of injectable erythropoietin-stimulating agents or, commonly known as ESAs, that are currently used in anemia of CKD.

This includes the potential to overcome the safety issues with ESA, treat anemia effectively in the presence of inflammation, where ESAs don't work so well and potentially eliminating the need for IV iron supplementation. We have study roxadustat in more than 1,100 subjects in our Phase I and II studies.

In addition to the potentially differentiated clinical profile versus ESA, roxadustat as an oral agent, has the potential to expand treatment except to CKD patients not on dialysis and who are not under the care of nephrologist. We and our corporate partners believe that most CKD patients are under the care of non-nephrologist physicians, such as endocrinologist, cardiologist and internist, until late stages of CKD develops.

It's estimated that 36% of diabetic and 20% of hypertensive CKD patients suffer from anemia.

Yet, the anemia goes untreated by their physicians because of the financial and logic -- than logistical barriers to prescribing ESA. If roxadustat enters the market, these non-nephrologist physicians will be able to prescribe roxadustat for anemia, while taking care of other concomitant medical problems like diabetes and hypertension, which are the leading causes of CKD anemia.

the dialysis and the non-dialysis pools.

In the non-dialysis Phase III studies, roxadustat is being compared to placebo. We aimed to confirm roxadustat to be superior on efficacy in anemia correction and non-inferior in safety relative to placebo.

In dialysis, roxadustat is compared to epoetin alfa. In this setting, we need to show roxadustat to be non-inferior to epo alfa for efficacy and for safety.

At the same time, our dialysis study pool is power to demonstrate safety superiority over ESA. As Tom mentioned, we and our partners are conducting 7 Phase III trials for meeting the requirements of U.S.

and European regulatory submission requirement.

FibroGen is directly responsible for 3 of these 7 studies. These are ANDES, which is otherwise known as Study 060 in non-dialysis patients; HIMALAYAS, Study 063 in incident dialysis patients; and SIERRAS, Study 064 in conversion dialysis study in stable patients.

Patient recruitment is going relatively smoothly for us and we expect to meet our base goal in reaching target enrollment in 2 of our 3 studies in the March/April 2016 timeframe, and to meet our stress goal of achieving enrollment completion in the third study by the end of 2015.

An independent data safety monitoring board has been a standing review body for roxadustat since 2009, initially appointed to oversee the Phase II studies and now all the Phase III roxadustat studies. The DSMB consists of independent experts who, in accordance with the DSMB charter, review the safety data periodically.

During its July 2015 review, the DSMB instructed us to continue the Phase III studies as planned.

I'll hand the call back to Tom now, and we'll be happy to address questions at the conclusion of the call.

Thank you, Peony, for your comments. I'll now ask Dr.

Frank Valone to make a few comments. Frank, please go ahead.

Thank you, Tom. Now, I'd like to return briefly to FG-3019 and our programs for treatment of fibrotic diseases and fibrotic cancers.

FG-3019 is a fully human monoclonal antibody that blocks connected tissue growth factor, or CTGF. A large body of scientific literature demonstrates that CTGF, central mediator of fibrosis.

We believe that FG-3019, by blocking the biologic activities of CTGF, may be a treatment for an array of fibrotic diseases and fibrotic cancers.

As Tom noted, we have seen profound evidence in animal models and clinical studies that FG-3019 has potential to inhibit and even reverse the progression of fibrosis. We completed an open-label dose finding study of FG-3019 in patients with idiopathic pulmonary fibrosis, or IPF.

That study showed that 35% of patients had improved or stable lung fibrosis after 48 weeks of treatment with FG-3019.

To our knowledge, no IPF clinical trial have shown improved lung fibrosis. This includes trials of the recently approved drugs pirfenidone and nintedanib.

Based on the encouraging data from our open label trial, we began Study 067, which is a randomized double-blind, placebo-controlled trial of FG-3019, that was originally designed to enroll 90 subjects with IPF. The study endpoints at week 48 includes change from baseline enforced vital capacity percent predicted, change from baseline and hold-on fibrosis measured by high resolution CT scans.

The goal has been to enroll the study by the end of 2015 and to have data approximately 1 year later.

[indiscernible] has slowed in the U.S. due to the approval and launch of pirfenidone and nintedanib for treatment of IPF.

In order to complete enrollment of the original study, we have opened up the trial in a number or countries where pirfenidone and nintedanib have not fully penetrated the markets. Countries we are now launching include Canada, Australia, South Africa, New Zealand and India.

In anticipation of a positive Phase II trial, we began to model several roots for product approval. Second line therapy after approving -- after failing approved treatment or first-line therapy when FG-3019 is combined with approved drug or possibly a single agent first-line therapy.

We already increased the size of our Phase II trial from 90 to 136 subjects in order to include subjects who had failed treatment with or pirfenidone or nintedanib. We are now considering expanding the trial further to include a group of patients who are receiving concurrent approved therapy.

We are still working on the details of the trial amendment. However, initial modeling indicates that we'll increase the size of the trial from 136 subjects to approximately 250 subjects in order to add a cohort of subjects receiving concurrent therapy, such as pirfenidone.

We will update further as our development decisions progress.

Duchenne muscular dystrophy has been a scientific and clinical interest of FibroGen for a number of years. DMD is characterized by extensive muscle fibrosis, and we consider DMD to be a fibrotic disease.

We and our collaborators have obtained compelling preclinical data that support FG-3019 in DMD.

This is a high-priority program for us because FG-3019 has a potential to treat all subjects with DMD regarding -- regardless of the underlying genetic defect. And thus, it's very different from most products in development that treat subsets of DMD patients with specific genetic defects.

As Tom noted earlier, we have recently received the FDA clearance on our IND, and our goal was to begin treating non-ambulatory DMD patients with FG-3019 in the fourth quarter of this year.

The study's primary endpoint is change in forced vital capacity. Boys with DMD have progressive loss of lung function due to increasing muscle weakness.

And for non-ambulatory boys, loss of lung function becomes a major health issue and ultimately results in full-time use of ventilators.

Other endpoints in this study are changes in arm function, which is critical to the boy's quality of life and direct measurement of arm muscle fibrosis by MRI. We are also measuring cardiac fibrosis by MRI, because heart failure due to cardiac fibrosis is becoming an increasing problem, as boys live longer due to improved pulmonary support.

We intend to expand our DMD program to include boys with DMD or ambulatory and we intend to meet with the FDA to discuss this plan next year. We are also evaluating FG-3019 as a potential therapy in certain cancers.

Elevated CTGF levels have been observed in diseases characterized by sustained production of extracellular matrix, or ECM.

Several cancers have prominent ECM components that may promote angiogenesis, metastasis and progressive disease. FibroGen has conducted both animal and proof-of-concept clinical studies of FG-3019 in pancreatic cancer.

In mouse models of pancreatic cancer, including the KPC mouse model, FG-3019 treatment demonstrate a reduction of tumor mass, slowing of metastases and improvement in survival. In an open-label Phase II trial of pancreatic cancer patients treated with FG-3019, plus chemotherapeutic agents gemcitabine and erlotinib, FG-3019 showed a dose-dependent improvement and 1 year survival rate.

We are currently conducting Study 069, which is a randomized trial in 40 patients with a locally advanced pancreatic cancer that cannot be removed surgically. In this study, patients are randomized to treatment with gemcitabine plus nab-paclitaxel or ABRAXANE, which is a standard of care chemotherapy regimen, administered alone or in combination with FG-3019.

The endpoint of this study is conversion of the tumor from an inoperable to an operable state. Complete surgical removal of the tumor is the best chance these patients have for cure of their cancer.

Additional study endpoints include tumor regression measured by CT scans, change in tumor metabolism, measured by PET scans and tumor markers, such as CA 19-9. We plan to submit an abstract of the initial study results or presentation at the ASCO GI meeting, which will be held in January of next year.

With this, I will turn back to Tom.

Thank you, Frank. Pat Cotroneo will now review the financial results from Q2 as well as some of the end of the year projections.

Thank you, Tom. As we announced in our press release today, total revenue for the quarter ended June 30 of 2015 was $120.6 million.

For the same period, operating expenses were $61.2 million and net income was 51. -- $57.1 million or $0.95 per basic share and $0.83 per diluted share.

Included in operating expenses for the quarter ended June 30, 2015 was an aggregate noncash portion totaling $9.4 million, of which $6.9 million was a result of stock-based compensation expense.

In terms of our cash balances, we had $408 million in cash, cash equivalents, investments and receivables, compared to $346.8 million at the end of 2014. Our investments consists primarily of 2- to 3-year investment grade corporate debt.

As Tom has mentioned, our cash balances at June 30, 2015, include 121 -- $120 million non-contingent license fee payment and the $15 million milestone payment received from AstraZeneca during the second quarter 2015.

In connection with the cost-sharing arrangement with AstraZeneca, FibroGen's total funding obligations for roxadustat outside of China are limited to $116.5 million, of which $83.4 million had been incurred and $33.1 million remained as of June 30, 2015.

Based on the current year budgets and enrollment forecast, FibroGen expects to reach this $116.5 million cap in Q4 2015, at which time Astellas and AstraZeneca will be responsible for funding roxadustat development in CKD through a launch for all territories outside of China.

Specifically, once the cap is reached, FibroGen will no longer be billed by AstraZeneca for 50% of AstraZeneca's development cost, resulting in lower expenses in our statement of operations.

In addition, once the cap is reached, roxadustat development expenses incurred by FibroGen will be fully offset by reimbursement. Looking to year end 2015, we continue to anticipate that our cash, cash equivalents, investments and receivables will be in excess of $325 million.

I'll now turn it back to Tom.

Thank you, Pat. Operator, we're now ready to start taking the question-and-answer session, if you can set that up, please.

Two questions, 1 on roxadustat. I know there's a particular focus on safety and of course, the DSMB analysis, can you just comment on to what degree or what additional information you could provide regarding risk of tumors, whereby you feel very confident other than just the carcinogenicity studies?

What kind of things were looked at with diligence? And in the DSMB analysis, are they specifically looking at that type of imbalance?

And the second question is just on pancreatic cancer, since there could be some data not too far away, just be a little more specific in terms of, I guess, where you are in that enrollment? How many patients do you think could be reported on at ASCO GI, things of that nature?

What would you be defining as good data to keep going?

Okay. So Michael, thank you for the comments and questions.

Let me do it this way, I will ask Frank to address the pancreatic question first, and then I'll come back and ask Peony to address the questions you asked about DSMB and tumors and so. So Frank, please go ahead.

Sure, Michael. Couple of things, we don't comment about patient enrollment, it's a lovely detail you might like, you find that might be misleading over time.

Enrollment is moving along smoothly. We have 3 centers now screening patients.

We will have a fair amount of data on patients who completed the study. As I indicate in my presentation, the primary endpoint is tumor resection.

But along the way, we can learn a lot about tumor responses by CAT scans, PET scans and tumor markers. So you have to go and pull the totality of data together.

We should expect to do it at the end of the year. As then as I said present in January, whether that's going to be enough data to make a decision to expand the trial or simply continue to just going, I won't know until December when I open up all those envelopes and see what we have.

But we feel quite confident we should continue the trial, we just don't know whether it will be expanded based on data we'll have in December.

Yes, so I might add 1 thing there, Michael, and that is these are patients that fail resection scoring before we start the study. And they go on a standard of care, gemcitabine and ABRAXANE and then 3019 in the randomized arm.

This means that if there's any patients that are resection eligible after 6 months, we've done something positive, and we've already seen that. And so, now it's just a question of how many patients do we need to see a difference, really taking on the question of standard of care, gemcitabine and ABRAXANE versus gem plus ABRAXANE plus 3019.

And so it's interesting times, I wish it would go faster, but methodologies are complex and getting people like Mayo and Georgetown on board has taken a lot of work just because they are learning a paradigm they haven't really been involved in before. So step-by-step, I think it's a good investment so far though.

So I'll stop there. Peony, do you want to address DSMB, I think Michael is asking about tumors, surveillance of tumors and so on?

Yes. So Michael, thank you for the question.

First of all, roxadustat corrects hemoglobin via a mechanism that is very different than ESA. And we understand the sensitivity about tumor risk in anemia therapy in patients who receive high doses of ESA as in the case of chemo-induced anemia historically.

And what we like to make it very clear is that our drug mechanism does not involve that excessive level of erythropoietin exposure. Furthermore, in the HIF biology, there is extensive work investigating the use of HIF-PHI for potential targeting treatment of cancers.

With that said, turning to your question regarding DSMB, our program has been under review by the DSMB since Phase II. During Phase II, we have not find any safety signal related to tumors.

During Phase III, our DSMB continue to review all SAE and tumors are part of the SAE reporting. And so we have not found any signals in the Phase III review as well.

One thing that we are very careful to avoid confounding by baseline factor is that we screened out patients with pre-existing renal cell carcinoma by ultrasound prior to entry in the study. So late stage patients who already have renal cell carcinoma, which is at higher prevalence in patients with end-stage kidney disease, would have been screened out of the studies.

Does that address your question?

It does. So there is a specific surveillance for it is, I think, what you're trying to...

So Michael, yes, the point here is that by doing the prescreen in ultrasound, we will be able to detect any incipient renal cell carcinoma during the study. We have not seen anything yet.

This is stuff that's surveilled pretty aggressively. But, at the same time, we've got a ways to go, but we have a way to have very clear readouts on that point.

So I was wondering if you could just comment on the regulatory timeline in China and sort of how long that submission process would take? And given the submission in 2016, when you'd get the approval there?

Okay, Kennen. I'm -- thank you for the question.

Let me confirm what you've asked about. You're saying, what are the regulatory timelines as we presently conceive China?

Yes, exactly. And just sort of timelines from submission to potential approval there.

Yes, so let me walk you through what we currently think. The study looks like it will start enrolling this fall.

We have been advised that about 4 weeks from now, we'll get to the point of having the -- what's called the chopped orders. So we have the authority to negotiate with IRBs [ph] and sites and contract for it and then start enrolling patients.

We have had, in preparation for this, several preceptor meetings, where we fly the clinical teams over to the U.S. and go through the data in great detail.

So we're pretty enthusiastic about moving ahead with the enrollments. And point of fact, we're doing 26-week study across the board, and then for safety cohort, we do 52 weeks.

At the point that we have 26 weeks fully enrolled, we are able under the green pathway rules to initiate a dialogue with the Beijing CFDA, which is our regulator at the level of CM&C activities. So the local regulator to CM&C in China and then national central regulator does drug evaluation.

Under the terms of the agreement it's a rolling review, we expect that the CM&C review will take somewhere between 9 and 12 months based on history and analogous situations. At the point we have the 52-week data, we're allowed to submit the drug review for CDE, the Center for Drug Evaluation, and make a formal assessment on the drug.

We believe this is somewhere between 5 and 7 months of time. So the way that AstraZeneca and we thought about this is that we're trying to make these 2 timelines meet up at the end of 2017, roughly December 2017, at the point that we would be authorized to make and ship drug.

Under the agreement with AstraZeneca, we are sending bulk drug shipments to AstraZeneca. It goes by railway to Wuxi City from Beijing, and we are paid 60 days later for the drug that we ship.

We expect that the terms of the Class 1.1 bio venture rules will require us to do somewhere between 2,000 and 3,000 patients in Phase IV. As a safety follow-up, the idea is on-market drug and a safety follow-up.

So China looks at these things. The administrator exclusivity rules are such that we would expect about a 5-year period from the time of approval until any competing products can conceptually be in the market.

However, this is not with -- this is not a judgment with respect to patents. This is simply a judgment with respect to administrative review in Class 1.1.

So we would expect that we'll have extensive patent protection after that time. So that's the basic idea, so you have an initial approval that is conditional, and it's conditional on the Phase IV safety review being adequate, clean in the eyes of the regulators.

And so when you think about it that we're aiming for December 2017 for the point we'd have the conditional approval and we would be thinking that something like 3 years -- 3 or 4 years later, we'd be done with the follow-up study and we would have the administrator exclusivity for 5 full years from December '17. Is that clear enough?

Yes, that definitely spells it out. And then maybe just one more question on 3019.

I think the sort of potential for a combination trial there is really a terrific one. And I was hoping maybe you could speak a little bit about the potential synergies that might arise in IPF?

And if there's one sort of mechanism of the approved products out there that potentially jumps out as being essentially more synergistic with the IPF patient?

So let me start by saying that we have been watching this year with the main strip submission on our exploratory study and looking at the peer review and the editorial comments and so on. And we are struck by the idea that the data seems pretty important to a lot of people.

And so in our hands, many years ago, 1997, we had an option to look at pirfenidone. And we declined it, as a matter of record, because we didn't feel that it could do what we wanted in fibrosis.

So we have a general view that the mechanism of pirfenidone is somewhat different than the one that arises from blockading 3019. And so we expect that it would be additive.

And so I would simply say that if we show that the HRCT methodology, which is really a time series of HRCT scoring, they take 1 millimeter thick cuts vertically and then horizontally in the patient, so they get incredibly accurate data for lung volume and for matrix and matrix deposition. If it turns out the data we saw in the exploratory study is replicated in this study, I think that there's a pretty compelling case for a combination therapy.

Recently, I think in part because of the review of the manuscript, just thinking about what's being said, we have taken on the idea of testing patients that have taken, for example, pirfenidone, in the U.S., because it's now been approved. So it'd be concomitant therapy kind of idea.

Let me stop here and have Frank explain what our thinking is on this, because we've spent a lot of time. Team -- Frank's team has really spent a lot of time on this past 45 days.

Tom has covered most of it I would say. We're actually looking at the question whether the combination of 3019 and pirfenidone or nintedanib would be synergistic.

There are -- they have different mechanism of action. So assuming they're additive rather than synergistic, we will look in preclinical models to see if we see signs of synergy, but it won't plan on that in our trial design as that might lead to trial failure.

I think we're pretty comfortable that additive makes sense and that we'd be looking at that type of trial design. At this point, I can't choose easily between pirfenidone and nintedanib.

They both operate pretty similarly, different toxicity but similar efficacy and they may just be additive at most. And part of our discussion internally is how we want to go so that we need to really do further modeling.

And as I said, we'll offer more comments once we got a chance to look at this further.

Yes, I mean, to-date, we've been thinking about, given the 90s studies -- this is end of the 90s range, that the total patients in the study would be something on the order of 250. We may go back and rethink that a little bit, but I think that's sort of a scale that we think we can handle and execute efficiently.

If we did that kind of number, we'd have to do 1 of the 2, pirfenidone instead of nintedanib or vice versa. And we just haven't quite got to the point yet of deciding about that.

Got it. And just to sort of follow-up on that, following the trial read-out done in IPF for next year, you'll have data from treatment-naive previously treated failures with pirfenidone and nintedanib and then also potentially the combination product.

So based on combination with 3019, so would the decision then be sort of how to design a registrational trial, and could you potentially go for head-to-head against approved products in treatment-naïve patients there?

So I think the answer here is that we clearly are trying to leverage our ability to develop a registrational strategy. And there is an option here depending on data to go head-to-head, but it depends a lot on data, and we don't yet know what to expect.

Frank, you want to add anything there?

We're setting up a program to have all possible options available to us in the end. What we do will be data-driven.

Maybe just one more on 3019. Just wondering if you can help maybe frame for us what type of FVC [ph] change you're looking for in non-ambulatory Phase II DMD trial?

Thank you, Terence. This is a topic that's been front and center here this year, and I think Frank knows a lot about it.

So I'm going to let him walk through.

Since we're still digging deep into that, I'm not sure I want to give you a number, because I'm looking at several models. The thing we're struck by is that in our trials, we see stable or better.

And so we have to decide -- still whether we look at FVC [ph] change or look at some other marker that may provide further differentiation from the currently approved drugs. So I don't have a firm answer to this question.

I think the one thing we can say, though, is that the KOL meetings we had, there's been an assertion that we should expect 5% decline in FVC percent predicted in these patients irrespective of ambulatory status. And that kind of number is sort of staggering when you try to get your arms wrapped around it, given what we've seen with IPF patients.

So I think that it's a target-rich environment here in terms of what may happen for us. I think Frank is getting at the question of what the measurement endpoint we're going to ultimately choose.

And that's -- but the concept here is that there's a really significant loss of lung function as the disease proceeds and...

What time course is that 5% over, Tom?

A year. It's actually more than 5% on average, but 5% to 8% per year.

What I saw was over a multiyear time period, this rate -- this slope continues. This was presented by the guy in Paris, the German guy.

I forgot his name, but...

Voigt [ph] .

Voigt [ph] , Thomas Voigt [ph] . And so -- and the other people there were concurring with this, I mean, that nobody disagreeing.

So you see really significant fibrotic impact on these patients. The concept here that is we originally noticed that there was a huge amount of fibronectin being made in DMD patients.

And as we dug into this, we realized that the dystrophin genetic defects were causing myofibroblast to exponentiating activity, so there is a massive amount of CTGF around, a lot of collagen around. And we just started digging away year-after-year working on this from 2006 until now.

22 papers that our collaborators have done. And it's very clearly the case that CTGF-mediated disease progression is certainly part of the picture here.

One of the factors we'll be paying a lot of attention to is there's some data that suggests CTGF dedifferentiates muscle in here, or in other words, they're turning muscle into fat. And if that's really true, then we might be in a situation where we're directly modifying the disease, which is sort of an ideal situation if it turns out that way for our purposes.

But we don't have enough data yet to be sure about that or some of the other stuff like the effect of reducing fibrosis on muscular strength, on endurance is very clear and so still to be explored some. But that's sort of the feel for it.

Does that help at all?

Just a few questions. Maybe first off, Tom, can you just help us better understand the -- your comment on the bundle and the determination there just in terms of its relative impacts?

I know this is potentially important from a commercial perspective. The second question -- yes, go ahead.

No, no, please go ahead and finish so I can figure out how to allocate.

And then maybe the second question, if you could comment on just sort of the next potential anemia indications, I know you mentioned a few in the prepared remarks, just in terms of the indications and then also timing initiation. I have 2 more questions, if you don't mind.

But just the frequency of DSMB evaluations in the global roxadustat studies and how many more could occur from here? And then the last question is, you mentioned other fibrotic cancers, fibrosis compounds.

What other cancers should we be thinking about beyond pancreatic cancer potential?

Okay. Let me do it this way, I'll start with the DSMB part and the bundle, and then we'll -- Peony can take anemia indications and Frank can do the fibrosis question.

With regard to the DSMB charter, we adhere to the charter so it's all prospective. And basically, there is preset prescribed activities each quarter.

This goes on for quite a while, in the future so much so that I don't know off the top of my head what is the point where it ends. It's been going on since 2009.

Several of the members have been on this committee since that time. So we just view this as part of the work environment.

And when I was talking to the Chair recently, at this point, in the study size, there is a lot of focus on whether there is any small imbalances and things that are completely unexpected, because this is a time period where if there's anything really off-the-wall happening, they try to cease on it. So it's -- I mean, pretty rigorous.

And I think people, while they believe that roxadustat is well tolerated, they also take this entire exercise very, very seriously. And so that's sort of the flavor of it.

With regard to the bundle, I know you weren't involved with us at the time, but from the FDA actions in 2011 that resulted in major changes in the use of EPO, until about 2014, I think the entire investment community was pretty bearish on anything involving anemia, in part because of the [indiscernible] failure, in part because the market seemed to be getting smaller and so on. And so it was a -- as a matter of course, it was assumed by everybody we'll be in the bundle.

And I think that you'll find that all the analysts that did any work assumed that we would be competing with biosimilar EPO for whatever paltry sum there is for some fraction of the bundle. We have had the benefit of a very excellent consulting advice, including people that used to run CMS, and they have a great deal of knowledge of how to think about these things.

And over time, the consensus built that the HIF side effects and our inflammation data, where we really don't see any dose increase requirements, when this lady gets the inflammation, hyporesponse seen in about 40% of dialysis patients in the U.S. where really high doses of EPO and iron are needed, there is a perception that we might trigger a view that some beneficiaries can benefit from this mechanism in a way that's not available for current standard of care.

And as we've thought about this over time, there have been various meetings and so on, and I think that at some juncture, this past quarter, I felt that it was inappropriate to allow the bundle assumption to just be unchallenged. And just -- we just sort of said, look, based on what we've heard so far, given that things can change and one doesn't know exactly what will happen and so on, and certainly, Amgen is out there with a lot of lobbying power, nevertheless, we felt that the better part of the argument was that we would be seeing separate reimbursement in terms of being new technology and the perception of the problems with the inflammatory anemias.

So what we really did was just simply modify disclosure to say that it appears that more likely than not, separate reimbursement, which is quite a bit different than the bundle assumptions. And obviously, many things need to be proven there.

So it's not a simple pathway to get that all figured out. But that -- Seamus, that's the idea.

Let me stop there on that particular issue and turn to the question of anemia indications. I think we've acknowledged in the call and remarks that we have been looking at other anemia indications with our partners.

Peony, you want to comment on that at all?

Yes. We always recognize that chronic kidney disease is a subset of many different conditions, which leads to anemia.

And given the mechanism of action of our drug as well as some of the preclinical work done in various other anemia, we do believe that our drug can be and -- have the potential to be of value in treating other anemia. And these -- this -- such work has been carried out over many years within FibroGen.

And you heard earlier from Tom about the potential to treat inflammatory anemia in patients with rheumatoid arthritis, lupus or inflammatory bowel disease. And we know that ESA becomes less effective when there's inflammation.

And we know that these are -- inflammatory anemias are not on ESA label. At the same time, as I mentioned, this is a subset of 7 or 8 different types of anemias, which we continues to -- we continue to evaluate and discuss with our partners.

At this time, we will -- we are not -- we will defer the answer about timing until further conversation.

Okay. So the last element of Seamus' question is for Frank, as you look at the various fibrotic cancers, what are you excited about beyond pancreatic?

Interesting. Seamus, I'll clarify our term fibrotic cancers, because it extends rather widely to leukemias, for instance, that require fibrotic stromal for the cell proliferation.

And 3019 in actually active in models of acute lymphoblastic leukemia, for instance. So there's actually a pretty wide range of catch that we could go after.

Ones that are particularly interesting at this point are glioblastomas, we have some pretty nice preclinical data, as well as squamous cell lung carcinomas. There's been internal discussion about some of the more difficult, orphan GI malignancy, such as cholangiocarcinomas at this point.

So I think we're really just working through this now. I have a new oncologist on the team who's really digging deep into our data, so he can really come up with a further development plan and hope to give you updates about this in the future.

First a quick question on 5200. Everything seems to be about China.

Is there no market, or have you not -- do you not think there's a market in the U.S. and Europe?

Is cadaveric enough there? Just talk a little bit about that.

So Tom, I think that's a very reasonable question. The way that this developed, I think that we, many years ago, started the pilot study.

If it's in its seventh year now, I guess, we work backwards and probably started preparing this study around 2006. So we didn't know for a while what we would be getting.

The occasion that caused the decisions around China was that in 2009 when we began to negotiate with the Beijing development authority, which is the entity that controls most of the property around Beijing, particularly in the science parts, we were informed that the -- there was a broad interest, most clearly personified by the Mayor of Beijing, who said that this is really important for China and we would like to see you bring this along with your anemia program as a second development platform. And given that China does not do cadaveric transplant to any degree of significance, I mean, there are some academic sites that do this, but it's not very much.

And as we learned about this, there are actually professional arenas where the government only allows blind people to work such as, for instance, being a masseuse. So this is a feature of Chinese life where the prevalence is quite a bit higher than in the U.S, for instance, or in Western Europe.

And so as we learned more about this and we learned about the reimbursement system for things made in China for China, we actually began to realize that the Chinese development opportunity was considerably more compelling than the alternatives in the U.S. and Europe and so on.

And so the general notion of development has been, for the past few years, let's do the pilot really pivotal study that's required in China, which we have been advised is around 75 patients. And then go from there to developing the necessary support for CE marking.

And so with that step, you move into certainly to Europe and Latin America and many other parts of the world, except the U.S. And with the U.S., where in the past, cadaveric transplants have dominated and only recently has it become apparent that there is a pretty serious problem with viral disease and the continued replacement of cadaveric transplant to the point where immunosuppressive lose their effectiveness.

We begin to see a subset population in the U.S. where there is a compelling unmet need.

And so this is still seen as less than 1/3 of the patients in the U.S. So it's not a dominant element, but it's significant enough that it gives us regulatory toehold.

So I think that's generally where we'll probably go in the U.S. is at some point or another, presuming success in China and the CE marking.

We would be looking at the patients that require transplants every 3 or 4 years, because pacification occurs due to the viral diseases and they're having problems with immunosuppressors, so clearly a population that needs an alternative. So that's been the thinking, and I'm sure it seems completely bass ackwards in some ways.

But given the circumstances of how things developed in FibroGen, there was a time when the offer in China was viewed as very, very compelling to us. And so we took it up, because in part, we wanted the anemia program to be enabled fully and which has happened.

So that's the background on that, if that's helpful.

Okay, so with that, I think we're completing the Q&A period. Thank you all for joining our call on this summer August afternoon.

I know you didn't -- any of you didn't have to do this, so we really appreciate your attendance and support. And we look forward to conversations at our next update.

Thank you very much.