Lumos Pharma Inc (LUMO) 2013 Q4 法說會逐字稿

Good day, Ladies and gentlemen and welcome to the NewLink Genetics fourth-quarter and full-year 2013 earnings conference call. At this time all participants are in a listen only mode. Later we will conduct a question-and-answer session and instructions will follow at that time.

As a reminder this conference call is being recorded. I will now read the forward-looking statements.

We would like to remind participants that matters discussed on this call contain forward-looking statements on the NewLink Genetics Corporation. I will now read the forward-looking statements disclaimer.

We would like to remind participants that matters discussed on this call contain forward-looking statements of NewLink Genetics Corporation that involve substantial risks and uncertainties. All statements other than the statements of historical fact discussed in this call are forward-looking statements within the meaning of the Private Securities Litigation Reform Act of 1995.

These forward-looking statements include amongst others statements about NewLink's financial guidance for 2014, enrollment in its clinical trials for product candidates based on NewLink's HyperAcute and IDO platform technologies, the timing of the release of clinical deals from ongoing clinical studies, its plans related to moving additional indications include clinical development, its plans for commercialization activities of any other product, candidates are approved for marketing, NewLink's future financial performance results, operations, cash position and sufficiency of capital resources to fund its operating requirements and any other statements other than statements of historical fact.

Actual results or events could differ materially from the plans, intentions and expectations disclosed in forward-looking statements that NewLink makes due to a number of important factors including those risks discussed in risk factors and elsewhere in NewLink's quarterly report on Form 10-Q for the period ended September 30, 2013, Form S-3 registration statements filed December 28, 2012, and in its other filings with the Securities and Exchange Commission. The forward-looking statements made during this earnings call represent NewLink's views as of the date of this earnings call. NewLink specifically disclaims any obligation to update or revise any forward-looking statements made on this call as a result of new information and future developments.

I would now like to hand the call over to your host Dr. Charles Link. You may begin.

Thank you for joining us today to discuss our 2013 operations and our expectations for 2014. With me today are Gordon Link, our Chief Financial Officer and Dr. Nicholas Vahanian, our President and Chief Medical Officer.

We have continued to advance and made solid progress in both of our cancer immunotherapy platforms, HyperAcute immunotherapies and IDO pathway inhibition programs. Let's start with the HyperAcute immunotherapy, which is designed to stimulate the patient's immune system to recognize and attack cancer cells.

For algenpantucel-L, our lead HyperAcute product, we completed enrollment in our pivotal Phase 3 IMPRESS trial for patients with surgically resected pancreatic cancer in September 2013. We recently announced that the independent data safety monitoring committee, or DSMC, completed the first interim analysis of this study.

The DSMC was scheduled to analyze the first interim analysis data during the first week of March after the 222nd event was reached in February. DSMC recommended for the IMPRESS study to continue as planned without any modifications. We were further reassured by the confirmation that there were no unexpected safety concerns.

A second interim analysis was planned upon reaching 333 patient events which is expected to occur sometime later this year with the final analysis planned, if needed, at 444 patient events. We are encouraged by the apparent lengthening of survival in the combined arms of this study because we believe the body of evidence in prior multi-institution trials of resected pancreatic cancer patients treated in the United States indicates that survival outcomes remain very poor in this unfavorable, unfortunate group of patients.

As we approach the second interim analysis we will continue our commercialization strategy and planning efforts including building the infrastructure needed to support an independent launch in the US market for algenpantucel-L as a treatment for patients with resected pancreatic cancer. For markets outside the US we intend to secure a partner for additional development where needed as well as for ex-US sales and marketing. We expect this timing to coincide with the interim analysis timeline for the IMPRESS trial.

We continue to actively enrolled patients in both our Phase 3 pillar study, which combines algenpantucel-L with the standard of care for patients with locally advanced pancreatic cancer and our Phase 2b/3 study of tergenpumatucel-L, our HyperAcute lung immunotherapy, which is being compared to docetaxel, an advanced non-small cell lung cancer. We have recently launched a Phase 2 study of dorgenmeltucel-L, our HyperAcute melanoma, in combination with the ipilimumab, Yervoy, for advanced melanoma patients. In addition, we recently launched a first-in-human Phase 1 clinical trial for HyperAcute renal immunotherapy in patients with metastatic renal cell cancer.

Looking to the future with our HyperAcute immunotherapies both algenpantucel-L and tergenpumatucel-L have demonstrated the potential to expand the benefits of chemotherapy by sensitizing the patient's immune system to chemotherapy. This chemo sensitization effect, which was presented at ASCO 2013, provides the opportunity for improved treatment outcomes for patients with metastatic disease receiving chemotherapy after prior exposures to HyperAcute immunotherapy.

The IDO pathway is regarded as a key immuno checkpoint target. Our IDO pathway inhibitor platform expanded and made significant clinical progress during this past year. IDO pathway inhibitors are designed to counteract a fundamental mechanism by which tumors evade immune mediated destruction.

Indoximod, our most advanced IDO pathway inhibitor, continues to advance patient enrollment in separate clinical studies indoximod in combination with standard of care for a variety of different indications including melanoma, pancreas, prostate, breast and brain cancers. We have recently initiated clinical trials with our second IDO pathway inhibitor drug, NLG919, in patients with advanced solid tumors.

We have found these two IDO pathway inhibitors to have differing mechanisms of action so they represent two distinct drugs which in fact have shown synergy with each other. We look forward to initiating a number of additional clinical trials with our IDO pathway inhibitors during 2014 including evaluating them in combination with other cancer immunotherapies such as other checkpoint inhibitors in cancer vaccines as well as chemotherapy and radiation.

In addition, we expect to report preliminary results for several of our ongoing clinical trials at appropriate scientific and medical forums throughout the year. In fact, two posters related to our IDO pathway inhibitors were accepted for presentation at the American Association for Cancer Research Annual Meeting being held April 5 through 9 in San Diego.

The first described the synergistic antitumor effects of combining checkpoint inhibition with anti-PD-L1 antibodies and our IDO pathway inhibitors, NLG919 and indoximod in the context of active immunotherapy. The second highlights the role of IDO pathway inhibitors in tumor immunotherapy more broadly. We'll talk more about this data once it has been presented.

We have seen significant interest from a number of parties for clinical collaborations as well as potential exclusive partnerships around our IDO program. Our business development strategy for this platform is to secure a global co-development partner in the near term.

We believe the best partner for NewLink is one that supports our long-term business strategy. Therefore, while it is important for this partnership to bring extensive resources for development and commercialization, it is critical that NewLink will continue to contribute scientifically, financially and strategically to the joint development plan going forward.

Of note, concerning our IDO inhibitor platform, in January of this year we held an event in San Francisco simultaneous with the Annual JPMorgan Healthcare Conference, in which we invited experts in the emerging field of checkpoint inhibition to speak based on their particular areas of expertise. I will briefly hand the call over to Dr. Nick Vahanian, our Chief Medical Officer, to provide a summary of the event.

Thank you, Chuck. Our expert panel for the IDO event included Dr. David Munn of Georgia Research University Cancer Center, one of the key pioneers to describe the immune function of IDO protein. Dr. Munn provided an overview of IDO pathway and its emerging role in cancer immunotherapy.

Dr. Samir Khleif, Director of Georgia Research University Cancer Center and a highly regarded leader in the field of PD-1 and PD-L1 discussed the synergy observed by combining indoximod, our IDO pathway inhibitor, with the inhibitors of PD-1, or PD-L1 function. Dr. [Ricky Holmgard], of the Memorial Sloan-Kettering Cancer Center, a researcher with considerable CTLA-4 experience, discussed the rationale and data supporting the inhibition of CTLA-4 function in combination with IDO inhibition.

The panelists discussed the fact that key immune checkpoint pathways, CTLA-4, IDO and PD-1 are interrelated and as a result combinatorial therapeutic interventions may prove more effective than single agent interventions. This was supported by multiple preclinical studies in which IDO pathway inhibition was shown to reduce local tumor mediated immunosuppression, providing potential for enhanced antitumor responses in combination with chemotherapy or radiation and in combinations with other immuno therapies such as cancer vaccines and other checkpoint inhibitors such as inhibitors of CTLA-4, PD-1, or PD-L1.

We know that CTLA-4 blockade has been shown to enhance tumor specific immuno responses that correlate with improved survival in melanoma tumors. As demonstrated with ipilimumab or Yervoy; however, the therapeutic benefit has been limited due to resistance mechanisms within the tumor microenvironment.

By combining anti-CTLA-4 and indoximod improved antitumor effects including reduced tumor growth and increased survival in multiple tumor models were observed. Importantly, this combination was well tolerated by mice with no weight loss or other clinical signs of acute or delayed toxicity or autoimmunity.

This activity was also observed with antibodies targeting PD-1. In addition, IDO blockade with indoximod enhanced the effects of PD-1 and PD-L1 blockade. Furthermore, in checkpoint inhibition resistant tumors employing the wholesale cancer vaccine product in combinations with checkpoint inhibitors, indoximod and ipilimumab allowed for effective immuno responses. This key concept is important since many human tumors remain resistant to checkpoint blockade.

In summary, our panel demonstrated that blocking the IDO pathway reduces immunosuppression and enhances immuno response and IDO inhibition is synergistic in combination with other immunotherapies, CTLA-4, PD-1, or PD-L1 cancer vaccines and, chemotherapy and radiation regimens. Indoximod is currently in multiple Phase 2 studies. Phase 1 and Phase 1 studies of indoximod demonstrated an excellent safety profile and it was well tolerated.

Noteworthy, we did observe that some patients developed autoimmunity but was reversible. Interestingly, the development of autoimmunity with checkpoint inhibitors has been regarded as a harbinger of drug activity with both ipilimumab and PD-1, PD-L1 blockade.

We have also demonstrated early evidence of the potentiation of chemotherapy with indoximod. In our Phase 1 trial, patients that progressed on prior chemotherapy were found to have objective responses when given indoximod with Taxotere. This included solid partial remissions in a patient with thymic carcinoma with previously failed Taxotere and a fifth line breast cancer patient.

Although this data is early we feel it represents important initial clinical activity of indoximod. These clinical observations are consistent with previously published data in nature and medicine about indoximod by George Prendergast, another member of our scientific advisory board.

Given that NewLink has two related yet distinct IDO pathway inhibitors, blocking immunosuppression and a portfolio cancer vaccine inducing immuno activation, we believe we are well-positioned to be a leader in the field of cancer immunotherapy. I will now have a call over to Gordon Link, our Chief Financial Officer, to review our financial results.

Thank you, Nick and thanks to all of you for joining us on our conference call this afternoon. We ended 2013 with approximately $61.5 million in cash and cash equivalents including the net proceeds from the sale of $17.5 million worth of shares under our ATM.

We expect to end 2014 with approximately $40 million in cash, cash equivalents and marketable securities including net proceeds from the sale of $27.7 million worth of shares under our ATM. Approximately $13.9 million remained unused under the ATM.

Research and development expense increased $4.9 million to $22.7 million in 2013 over 2012 primarily due to expanded clinical trial activities for our IDO pathway inhibitor programs, NLG919 and indoximod. General and administration expense increased $2.4 million to $9.5 million in 2013 over 2012 primarily due to increased personnel related expenses as we prepare to launch algenpantucel-L should it be approved by the FDA.

We currently expect our cash resources to allow us to fund operations through 2015. The loss for 2013 was $31.2 million increasing from $23.3 million for the year ended December 31, 2012.

Comparison of our per share losses is not meaningful as our weighted average common share count changed due to shares issued in NewLink's follow-on offering in February, 2013. With that I would like to turn the mic back over to Dr. Link.

So to summarize, in 2013 and early 2014 we advanced all of our major development programs and look forward to the following milestones for the remaining of the year. Let's start at the HyperAcute immunotherapy platform.

As we mentioned earlier in the call based on the first interim analysis we will continue the algenpantucel-L Phase 3 IMPRESS study for resected pancreatic cancer with the second interim analysis expected upon the 333rd event later this year. We plan to continue accrual in our algenpantucel-L HyperAcute pancreas Phase 3 pillar study for locally advanced pancreas cancer with our preliminary data for presentation anticipated in early 2015.

We plan to continue patient accrual in are tergenpumatucel-L HyperAcute lung study and Phase 2b/3 in advanced non-small cell lung cancer. In 2014 we launched a Phase 2 study with dorgenmeltucel-L HyperAcute melanoma in combination with ipilimumab for advanced melanoma patients and we intend to complete enrollment in our recently initiated Phase 1 HyperAcute renal study for metastatic renal cancer. We also plan to initiate combination immunotherapy trials in 2014, evaluating the activity of our HyperAcute immunotherapies in combination with our IDO pathway inhibitors.

Turning to the IDO pathway inhibitor program, we intend to be very busy with clinical efforts to the program for that program as well. Let's start with indoximod. Late this year or early 2015 we anticipate presenting clinical data for our ongoing Phase 2 study of indoximod plus docetaxel in breast cancer patients.

We will continue to advance our programs for our ongoing Phase 2 study of indoximod plus PROVENGE in metastatic castrate resistant prostate cancer, and our Phase 1/2 clinical study of indoximod plus temozolomide in advanced brain tumors. Furthermore, we intend to make rapid progress in our Phase 1 studies of indoximod in combination with ipilimumab for advanced melanoma and indoximod in combination with gemcitabine and albumin-bound paclitaxel in metastatic pancreatic cancer.

Shifting gears and to NLG919, we have the following milestones planned. We intend to present preliminary data for our ongoing Phase 1 study of NLG919 in advanced solid tumors late this year or early 2015. We plan to initiate multiple Phase 2 studies of NLG919 plus indoximod plus HyperAcute immunotherapy in solid tumors and additional studies evaluating NLG919 in combination with chemotherapy and/or other checkpoint inhibitors.

As you see there are no shortage of clinical activity this year and we look forward to keeping you apprised of our progress in the weeks and months to come. With that we will open up for questions.

(Operator Instructions). Salveen Richter, Cannaccord.

This is Andrew on the line for Salveen, and thanks for taking my question. Just to start with could you remind us of the overall survival benefit you will need on the second or third analyses to stop it for efficacy and can you just kind of give us some thought about how you're thinking about the likelihood of hitting that benefit given the Board's decision on the first analysis. Thanks.

We still, as we previously said, the first interim analysis was the statistical high bar and we knew that. The main reason being that a number of the patients that were enrolled, about half the patients enrolled during the last year and three months of the study had not had very long follow-up to be able to create enough differentiation for that segment of the population to achieve that statistical high bar and that P-value of 0.004.

The second interim analysis, which we anticipate will occur later this year, will have approximately, it will require approximately a 30% improvement between the two arms based on log rank analysis and the P-value for that is 0.019, what we feel a much more realistic bar in terms of activity to achieve. And we think that there is a significant potential for that interval analysis.

However, of course, the study is designed to detect a 20% difference in overall survival based on log rank analysis comparison to the two arms of the study at the final analysis, which we believe will occur during 2015 and will be triggered when there are 444 patient events. That final analysis actually requires a P-value of 0.043.

If I could emphasize on that one more time, an important point which is the 30% approximate improvement as Chuck mentioned is log link analysis without actually knowing the curves, how the curves separate. That's a projection of the estimated survival between two groups and how the curves separate and how the expected benefit between the two curves, so it's an approximation of around 30% benefit.

Okay. That's helpful, thank you. And then just a quick follow-up on the IDO pathway. When you are thinking about partnering are you talking to people about partnering both drugs together, or are these completely separate?

So the answer is it depends upon the interest of the partner. We could partner one asset separate. We could partner both assets together, or the second asset by itself.

So it really depends on the partnership and what the partner is interested in and how well their clinical development plan fits with our clinical development plan for what we think is a terrific class of checkpoint inhibitors. I think it's quite interesting that we found that IDO inhibition seems to be right in the center of the pathways that relate to CTLA-4 and PD-1, so it seems to be a very central and important target.

In other words, Andrew, we are pursuing all opportunities between one compound or the second compound. We're open to discussions and we are evaluating all opportunities.

Okay. Thank you very much.

Jason Kantor, Credit Suisse.

Good afternoon, this is Jeremiah for Jason. In terms of the trajectory for the second interim analysis at 333 events you mentioned that you expect that sometime this year.

How confident are you in terms of the trajectory for that readout? Do you expect roughly within a certain period or a certain timeframe, or is that range kind of broad still?

Well, before we reemphasize the later this year comment about the second interim analysis or 333 events, I would like to reemphasize a point that in our Phase 2 the benefit projected from the HyperAcute vaccines went up at one year at 37%, 38%, the second year 58% and then final third-year analysis showed over 100% benefit. So there is an increased benefit from immunotherapies in general and I believe, we believe, our HyperAcute vaccines have demonstrated that as well. Keeping that in mind having the 222nd event in February, our projections at this time is towards end of 2014 will be 333rd events.

You spoke to that in terms of you saw that increasing benefit over time for in your Phase 2 study and that's something it seems pretty characteristic of immunotherapies where you have a Kaplan-Meier curve start to level off as time goes on. Would you, just hypothetically speaking, would you expect to see that before the second interim analysis, just kind of theoretically, or is that something you might see more towards the final analysis?

I think that some of those events are beginning to accrue in the data currently. If you think about the first patients that were enrolled in early 2010, those patients are going to have three years or more follow-up here in the not-too-distant future.

And so if you look at the survival curve from Phase 2, we had a flattening that occurred between years two and years three in that survival curve. The tail effect, if you will, has been a uniform part of immunotherapy that has been seen with PROVENGE, it's been seen with ipilimumab, it's been seen with PD-1 inhibitors.

And we believe that we saw a similar tail phenomena in our Phase 2 testing in a variety of different diseases with HyperAcute technologies. So we do believe that the pattern of response is going to be similar with regard to that.

So it's very distinct than what you see with chemotherapy or small molecule inhibitors where a lot of the beneficial effects are dominant early and fade later. Here you see sort of the opposite pattern that seems to be a general principle for immunotherapy that we think will also hold true for the HyperAcute program.

This is a follow-up. Is it possible at all to predict when that might happen, or is that just something you just -- because you are saying there is sort of a range on patients who are enrolled but is that just something you can't really predict beforehand, it's more at the near the after the analysis is done or retrospectively?

No, I think that you will be able to see the curves in real time once we have access to the data but obviously we are blinded to the data currently. We continue to believe that the control group performance is not going to vary dramatically than what has been seen previously before in large multinational, multi-institution trials in the United States and that US trials have a particular pattern of recurrence and death that have been reproducible over several decades.

It is true that the advent and use of Abraxane and FOLFIRINOX may be adding a month or two, or three to those survival curves. But we think that the beneficial effects of those new agents that have limited ability in metastatic disease and second line recurrent metastatic disease will also have limited benefit and therefore we find that encouraging.

We think that it is very important to distinguish the clinical pathological patterns of pancreatic cancer from other types of malignancy that can have tremendous variation and outcome. The outcome in pancreatic cancer, unfortunately, for patients has been one of the toughest ones that we have to deal with. And I can tell you that as having been an oncologist for the last 30 years it's one of the toughest group of patients that we ever meet in the clinic.

Think your particular questions.

Mara Goldstein, Fitzgerald.

And I just wanted to ask a follow-up on that issue, on the question around control and how the control arm might be acting. Given that the events, the 222nd event happened in February, is it your expectation then that the control arm is performing as you would have figured it in this statistical plan at this point in time based on when you know patients were enrolled in the trial?

So we designed a statistical plan that would easily tolerate a control arm in the low 20s. And we did that purposely even though we know historically in the United States the outcome for instance of the RTOG-9704 trial was 18.6 months if you include all the patients in that trial.

So our view remains the same as we've had all along, which is there may be some benefit from these new chemotherapies that have been approved but the benefit we believe from those treatments will be modest. And we don't believe that there's any fundamental change that has occurred in the United States that is suddenly going to jump the survival of pancreatic cancer patients in the control arm by five or six months.

We don't believe that. We obviously don't have access to the data, we are blinded to it, but we have been reviewing and thinking about this issue a lot and reviewing other data a lot.

Okay and if I could just ask a follow-up question on the question of partnering -- you said that you would like to partner but you have some requirements around that, some would be participation. Can you talk about what that would mean in terms of personnel involvement as well as financial involvement?

Yes, we very much feel that we have been in the checkpoint inhibitor field for now more than 10 years, when we first merged the IDO technology into our Company by taking over and merging in a privately held company in Philadelphia that was run by George Prendergast that had some intellectual property related to IDO. We subsequently then licensed the bulk of the intellectual property from Georgia Regents University, David Munn and Andrew Miller's group, that did a lot of the pioneering work on IDO as a target.

Our view is that with that team we have a very good sense of what we think the logical combinations of checkpoint inhibitors are, the logical sequencing of vaccines followed by checkpoint inhibitors. We sense that some large companies have not spent as much time thinking about the importance of driving effector T cell function before you do checkpoint inhibitor blockade. The reasoning is if you do checkpoint inhibitor blockade and there aren't the right type of T cells around in the environment to create effective antitumor responses, even if you release them and allow them to proliferate they might not have the right targeting.

So if you do vaccinology prior to that and drive the immune response and then do checkpoint blockade, all of the preclinical work suggests that in some instances you can double the effectiveness of the checkpoint blockade. More importantly, and this was actually some data recently published by [Ricky Holmgard] and one of the reasons we had her talk at the JPMorgan event.

Some tumors seem to be completely resistant to single checkpoint blockade inhibitors or multiple checkpoint blockade inhibitors in combination. But they found in some of their experiments that if you give a whole cell vaccine prior to the checkpoint inhibitors you can revert the phenotype and make the tumors sensitive again.

I think that's going to be a general principle. And of course NewLink was founded on that fundamental principle of believing that driving the effector immune response followed by checkpoint inhibitor blockade was the right combinations ultimately in terms of how to employ immunotherapy.

All right. Thanks, I appreciate it.

Biren Amin, Jefferies.

I guess I'll start with HyperAcute pancreas. On the HyperAcute pancreas interim why shouldn't we assume that the control arm would be living beyond the low 20s?

So for example could we make the assumption that the control arm is living at 24 or 25 months, and if so what does that do to your stats assumptions? Thanks.

I'll start by referencing a recent study that was published by Johns Hopkins Group which demonstrated that for the last three decades going all the way back to the 1980s, 1990s and all the way up to 2011, the survival expectancy of pancreatic cancer was 19.2 months. In all three decades survival did not change in the United States.

Looking at the RTOG study, which was the largest pancreatic cancer study completed prior to ours and resected patients, the median survival was 18.6 months. The benefit of GEM/Abraxane combination in metastatic studying up front is 1.7 months. Assuming all of our patients receive GEM/Abraxane follow-up in the salvage setting after recurrence and assuming that even in the recurrent settings they are going to benefit as much as they would in the upfront settings, that would move the needle from 18, 19 months to low 20s at best.

And there is no between [size pool] (inaudible) and similar benefit and again in a metastatic setting. The benefits are limited in pancreatic cancer for the last few decades.

Considering that it is our expectations, it is our belief that in our study today we don't have any reason to believe that median survival for these patients will be more than low 20s. Nevertheless, our study even though expectations were 18, 19 months, study is designed in the low 20s to be able to -- is powered around that to be able to capture the difference around 20% in survival for the final analysis.

So a statistical plan has been prepared to capture the difference around 20%, as little as 20%, with control group coming in the low 20s. We believe that is the reason for our confidence for the statistical plan for the study. Chuck, do you want to add to those?

Yes, if I get your question, I think even if the study control arm comes in at 23, or 24 or 25 months, there is certainly still the opportunity as such this study could be positive at the final analysis given where the projected overall means for the study are currently. I know that one issue that came up that somebody had questioned was that there was an enrollment criteria for the Phase 3 protocol that patients have expectations of six months of survival and as I understand it the RTOG study had an expectation for enrollment of three months of survival.

We don't believe that entry criteria has any material difference between the two studies. Specifically, the decisions to go to surgery are made at most of the universities by teams of surgical and medical oncologists, sometimes in concert with radiation oncologists.

And if they don't, in general, believe that the patient has an opportunity to live at least six months generally they wouldn't be taking the surgery. And that's very similar to the decisions that are made by many of the similar types of teams, at many of the same universities that participate in the RTOG-9704 study in our study.

So that criteria we don't believe is selecting for somehow a different group of patients that were present in the last national trial. I think one of the things that would argue that point to be true is when we published September of last year the initial demographic characteristics, which we hope to put out in more detail at ASCO, that the patient populations from a US only multi-institution trial at more than 70 sites looks very similar between the RTOG-9704 study and the NewLink IMPRESS trial.

So I think if the patient demographics are similar, and the stages the patients entering this trial are similar, they are at similar institutions and it's a similar large study, we think that it's going to reflect what that US population looks like. The fact that the Johns Hopkins data they published overlaps with two years during the period when NewLink was enrolling patients on the Phase 3, they show up between 2000 and 2011, 1,100 US patients a median survival of 19 months, I think that Hopkins is prototypical for the type of university tertiary referral centers that are receiving and treating patients under our Phase 3 IMPRESS trial.

I would like to add one more point. Again I want to reemphasize that our expectations being around low 20s, 21, 22 months for the control arm for the reasons that we just went over, and I will use the number that you used, maybe 23, 24 months.

Even if you go higher with that for some reason if the control arm comes at those levels even though we do not expect it nevertheless we have as you remember we have demonstrated in the Phase 2 study we had three patients with CRs that had recurred after vaccine and the salvage treatment had complete responses. So the chemo sensitization effect that we have demonstrated with our tergenpumatucel and algenpantucelble pancreatic cancer vaccine that we believe further benefits the vaccine.

So in this trial if there was a high benefit to the control arm that impact a little bit more in the study arm. So we believe for some reason that if we are miscalculating, you sort of think low 20s and the number comes 23, 24 months, the impact will be even larger in the study arm for that reason.

Okay, maybe if I could get a follow-up also on HyperAcute. What's the baseline patients that are R0 versus R1 in the study? I think you've disclose some baseline characteristics last fall but that wasn't one of the criteria that you had disclosed.

And then also I think the radiation use in IMPRESS is at investigator's discretion and so I wanted to maybe if you could share the level of radiation used in the IMPRESS trial if you could potentially share that information. Then lastly I guess on the timing of the second and on the final, why given that all the patients are enrolled in the study, why wouldn't we expect to see an acceleration of events now that these patients have been enrolled for quite some time? Thanks.

Sure. So I'll start with your first question.

The patient characteristics that we put forward right after the enrollment completed September was important prognostic indicators that impact survival which was also our stratification factors, mainly nodal status and CA19-9 levels. When we started the study in May 2010 when the first patient was enrolled and prior to that designing of this study we offered very careful consideration and talking to all the key opinion leaders and reviewing all the literature we've decided that nodal status and CA19-9 levels are the most important prognostic indicators.

Based on that when the study was completed we released a criteria for nodal status and CA19-9 levels in our Phase 3 study to demonstrate that our patient characteristics were very similar to expected US population where our nodal status was 70% in our IMPRESS trial, where RTOG-9704 was 68% in both trials, CA19-9 and nodal levels, were 9%. But that was our reasoning behind why we chose those characteristics. To your question about further characteristics in the upcoming meetings we will put out more discretion about the patient characteristics in the IMPRESS trial at the appropriate time in the meetings.

I do think that there was a little bit of a teaching point in terms of R0 versus R1 resections. That's important, we believe.

If you look at single institution trials that patients who have smaller tumors or node negative tumors, or only one node positive with adequate node sampling, which usually means 10 or more nodes that have been sampled, in subsets of patients that have small tumors, generally less than 2 centimeters, and 0 or 1 node positive, those patients are the ones that have by far and away the best prognosis. And that's true in any population of resected pancreatic cancer patients from almost any institution.

In that subset of patients R0 versus R1 in single institutions has been demonstrated to be a very important prognostic factor. That's been demonstrated in a study recently at St. Louis University, I believe it was. There was a prior publication that came out of a single-institution study from Johns Hopkins.

And so in that setting R0 versus R1 is important. If patients are node positive with two or more nodes, if there is adequate sampling,, or probably true if there's one more nodes in inadequate sampling, or larger tumors that are greater than 2 centimeters or 2.5 centimeters in size, there does not seem to be an independent impact, statistically significant impact of R0 versus R1. Since the majority of the patients are node positive or have larger tumors then the impact of R0 versus R1 in single institution studies is limited to that smaller subset of the patients, which is less than half the patients.

In multi-institution trials, surprisingly, people have studied this effect including the RTQG-9704 study that was published and found that there wasn't independent prognostic effect, very limited value in R0 versus R1 comparisons. The reason that the experts believe that's the case, at least what they've told us, is that in multi-institution trials you have the problem that different institutions define R0 versus R1 differently.

So some institutions do it on a certain number of millimeters from the edge, some right on the edge, some 5 millimeters from the edge. It varies.

It is also the case that to do that pathology it has to be done very carefully and very completely to truly define where all the true possible tumor margins are after the resection. And so when you look at multi-institution studies, and this was true both in our Phase 2 study where we compared R0 versus R1 outcome and found no statistical difference in survival in our Phase 2, and it was also true in the RTQG-9704 where that did not turn out to be a significant variable.

So we anticipate, and our data will be able to analyze this of course, that there isn't going to turn out to be much different in our large 70-plus institution trial between R0 and R1, in part because it's dominated by the fact that the node positive and larger tumor patients are the majority of the patients. So it's only a smaller subset that falls into that category but also in part because of variability across the 70 institutions for how thoroughly or completely or how they define R0 versus R1 resection.

And if I can comment on your last question, which was about timing for the second and final analysis. If you consider now 222 events took close to four years starting in May 2010, we in fact accept your premise that there will be some acceleration of events occurring because next 111 events we are saying that is going to happen towards the end of 2015 -- 2014 -- which is 9, 10 months from now, 8 to 10 months from now.

So that in fact considers that. But you have to consider the balance between late impact, or exaggerated impact or late impact of immunotherapy as the time progresses. So it's got to be the balance between patients benefiting more and more from HyperAcute immunotherapy and acceleration of events because of a higher number of patients in the pool.

That's why we are projecting towards end of 2014. Does that answer your question?

Yes, great. Thanks for taking my questions.

Stephen Willey, Stifel.

Maybe just a follow-up on that last point. It looks like the first interim was triggered when the median study population, the median OS of the study population was in the 26-month ballpark, would you anticipate that median study population OS to hold up as we go forward?

We can only make projections about this but our impression is that the overall median is in the high 20s, somewhere. So I can't tell you exactly 25 months or 26 months, or 27 months, or 28 months, but we believe it's in the high 20s.

We'll have to see what happens over time in terms of that overall median. Obviously we will begin to get a solid sense of that when more and more patients get out further on the trial to see if there's a shift in that median.

There has been no appreciable difference in entry criteria throughout the conduct of the trial. So we believe the patients that were enrolled at the beginning of the trial are very similar to the patients enrolled at the latter stages of the trial.

We did make one minor trial modification during the conduct of the trial where we allowed patients to be out to 9 weeks, out to 10 weeks after surgery before enrollment because there were some patients who wanted to get into the trial that just weren't recovering from surgery fast enough. Those patients tend to be a little bit sicker clinically but the surgeons thought that they would be good candidates and so we modified the criteria from 8 weeks to 10 weeks after surgery to still allow them to be permitted into the trial. And that modification was done I think in early 20 --

In 2012.

Early 2012. But really we believe the patient characteristics are going to be fairly uniform across the study and look very much like a US population and not like a European-style population.

Okay. You kind of talked about the flattening of the OS curves that occurred between two and three years in the Phase 2 study and as we start to get into that period whereby you get a greater percentage of patients in that range, I'm just trying to think about what kind of obligation DSMB may have beyond just taking a specified event driven look at the data. And if by chance I think there's some extension there, is there an opportunity for DSMB to potentially step in in absence of actually hitting one of these predefined event numbers?

As you know, Steve, their obligation is to the patients. And their obligation is to do what is clinically best for the patient. So if they discerned that there was a substantive benefit they are always at -- they always have the right to stop the study when they see fit if that were the case.

The data it currently is planned to be at -- the next round of data would be aggregated and presented to them would be when they are at 333 events barring some other request that we didn't know about from DSMC. But they have the right to ask for data or to do modifications or things that they see fit.

Obviously, this is what we believe to be a very well defined, rigorously designed trial that will really, I think, prove, hopefully, the effectiveness and that we are trying to do. But they obviously are in the responsibility, in the driver seat for the patients.

Steve, if you also consider the fact that all patients are already enrolled in the study and nobody is being randomized in the placebo, so there's that kind of pressure is not present for the DSMB to make any, or any aggressive decisions about when they end the trial. I think statistical endpoints are well defined.

It's on their SPA, as you know. So the endpoint considering it would be currently based on schedule and that would be my expectations.

Okay, then I have a question you've been kind of asked a lot too, was I guess the rationale to structure an interim when only half of the events have occurred. And I guess in the context of this kind of statistical disadvantage that we know that immunotherapy tends to be from kind of a KM statistical perspective. Just wondering if you can provide any color around that strategy?

Sure. The strategy that we have employed it is early look. We expected that.

But nevertheless the fact that we lost very little alpha, the final analysis will be 0.043, so with this single study under SPA we lost very little alpha. Because of that -- and we also learned that there is no safety concerns for the product and there was no modifications suggested in the first interim analysis.

And when we started this over four years ago with the statistical plan it seemed like a reasonable plan getting an early analysis even though all along as we have communicated this is a very very high bar with a P-value of 0.004 and we have emphasized our low expectations for the first interim analysis. Nevertheless it was put in place because it was also cost because the bar was so high it cost very little Alpha.

The second interim analysis, on the other hand, with a P-value of 0.019, is a reasonable expectation. And we have confidence in the final analysis with the power of 80 as the study was designed.

Steve, when we look back at the Phase 2 data I think that what we were struck by and what we didn't really fully understand, as you know, the Phase 2 data didn't have a no-treatment arm so the Phase 2 data was limited. But one thing that we know was true is that in the high-dose group at one year, out of the 26 patients, only a single patient had died and that really exceeded any expectation that experts in the field had for what would happen in terms of one-year survival.

And that one patient, when we looked at that patient, had a 12 centimeter tumor, a fairly enormous tumor from a pancreatic surgery perspective. So we felt that that single fact, the survival at one year, was a very strong efficacy signal for the trial in the high-dose group.

And we did believe that we'd seen evidence of dose responsiveness in both lung cancer in pancreatic cancer that then led us to do this 12-month treatment plan. And so since we didn't lose much alpha we thought why not take an early look at this? If there was more dose responsiveness and more duration responsiveness in the clinical effect maybe there could be something of that magnitude, so that's in part why we did it.

Okay. And then just shifting quickly to IDO. How set in stone are some of these proposed trials just with respect to timing and initiation?

And I ask the question just in context of the color that you provided with respect to exclusivity around the potential partnership. And I know oftentimes when these exclusive partners are brought on board they tend to have a bit of a development agenda and so I'm just curious as to whether or not the plan that you have outlined are definitive plans for 2014, or are those all kind of subject to change if anything happens on the BD front?

Yes, so as always, as you are a developmental stage company, we are going to follow our nose in the clinic. We consider ourselves to be fairly good at looking at clinical activity, clinical scenarios and trying to develop trials and things around them.

Obviously the clinical development team has been doing that kind of work for 30 years. So we feel that as we see different activity or different occurrences hopefully in the different varying Phase 2 trials the intensity and speed with which those will transition or grow to larger trials or be transformed into registration trials will be very dependent upon what kind of clinical activity and observations we have in the clinic.

And so it's the end of observations in the clinic that come out of Phase 2, obviously we will try to let people know about those sooner rather than later. In terms of limitations of a large corporate partner and limited team development of the drugs, I think that we would feel very strongly that on the trials that we are launching that we think have the strongest preclinical or clinical evidence to support them, that we would be reasonably headstrong about making sure that those studies went through. Now as you know everything is in negotiation and we would make rational good decisions with that partner because we would want a long, good long-term excellent partnering relationship.

Okay. And then just lastly, just one housekeeping question. Maybe for Gordon if you can just give us what the cost basis of the ATM withdrawals were in 4Q and then also year-to-date and how much of that ATM is remaining? Thank you.

I guess maybe you missed that piece. We have just under $14 million left on the ATM and we sold about $27.7 million worth of shares on the ATM in the first part of this year and the remainder in the end of last year. So altogether we sold about $45 million under the ATM out of a $60 million selling agreement.

But is there a cost basis that you can provide us with?

Well, it's no more than 4%. That's what the selling agreement --

No, cost basis on a per share basis. So you sold --

Yes, there's really no problem with that. Its range is about $25 a share.

Okay. Thank you, guys.

Thank you. And that does conclude our time for Q&A today. I would now like to hand the call back over to Charles Link for any closing remarks.

Well, thanks everybody for dialing in today and we will continue to work hard and focus on developing these immunotherapy platforms. And many thanks to all the physicians, investigators and all the clinical patients who are participating in our clinical trials who without their support we wouldn't go anywhere. Thank you very much.

Thank you ladies and gentlemen for participating in today's conference. This does conclude today's program.

You may all disconnect. Have a great day, everyone.