Lisata Therapeutics Inc (LSTA)

[andy e]

NeoStem's NBS10 And Data That Matters To Regulators
Nov. 20, 2014 7:30 AM ET | 24 comments | About: NeoStem, Inc (NBS), Includes: BAX
Subscribers to SA PRO had an early look at this article. Learn more about PRO »
Disclosure: The author is long NBS.

Summary

Recent positive interim data was reported on NeoStem's previously blinded Ph II study, which the investment public has misinterpreted to be negative.
In every endpoint important to regulators, NBS10 showed promising signals of efficacy.
Partnership for NBS-10 seems likely, and would front the cost of a pivotal Ph III trial.
On Monday evening, NeoStem (NASDAQ:NBS) reported interim data on their double-blind placebo-controlled 161-patient Ph II study testing NBS10 in patients that had experienced acute myocardial infarction (AMI). With the exception of a noticeable difference in perfusion between groups, as measured by SPECT, every other clinical endpoint, especially those required by regulators to demonstrate effectiveness of therapy (namely overall survival, serious adverse events, MACE and LV ejection fraction), showed a positive difference between groups in favor of the NBS10 arm.

This article was sent to 5,075 people who get email alerts on NBS.
Get email alerts on NBS »
Specifically, the following was reported:

A statistically significant mortality benefit (p<0.05) in patients treated with NBS10 (also known as AMR-001) as compared to the placebo group; there were no deaths in the treatment group.
A statistically significant dose-dependent reduction in SAEs (p<0.05).
Observation of a dose-dependent numerical decrease in MACE. MACE occurred in 14% of control subjects, in 17% of subjects who received less than 14 million CD34 cells, in 10% of subjects who received greater than 14 million CD34 cells, and in 7% of subjects who received greater than 20 million CD34 cells.
When correcting for the time to stent implantation in all subjects, patients treated with CD34 cells were seen to have a statistically significant dose-dependent improvement in their ejection fraction (p<0.05). Independent from time to stent implantation, a statistically significant improvement in ejection fraction (p<0.05) for patients treated with a dose of greater than 20 million CD34 cells compared to placebo was observed.
No meaningful difference in perfusion, as evidenced by SPECT imaging, between the treatment and the control group from baseline to six months in resting total severity score (RTSS), suggesting this may not be a future suitable tool to assess NBS10, which is consistent with U.S. Food and Drug Administration (FDA) guidance that mortality and MACE are the appropriate approvable endpoints to determine efficacy of a cellular therapy for cardiac disease as opposed to imaging endpoints.
A particular handicap in favor of control group also surfaced:

Prior to trial enrollment, patients subsequently randomized to the treatment group had experienced a significantly longer time to stent implantation after the onset of symptoms (931 minutes) compared to subjects subsequently randomized to the control group (569 minutes). This longer interval to reperfusion would ordinarily be expected to be associated with worse clinical outcomes.

Yet despite this obstacle, the treatment group showed a statistically significant 40% difference in MACE for those administered over 14mil cells, and a 100% difference in MACE for those administered over 20mil cells. These two groups likely account for the majority of patients enrolled in the treatment arm (due to the percentage difference being statistically significant).

In the conference call the following morning, it was revealed that the number of stem cells administered to patients reached a maximum of 100mil cells. Given that there was a clear, statistically significant correlation between those dosed above 20mil cells and a 100% difference in MACE (and it was sequential: over 14mil fared better than less than 14mil, and over 20mil better than 14-20mil, further lowering the probability it was not dose dependent), as well as a stat sig improvement in ejection fraction, regardless of time to stent implantation - 20mil cells will likely become the new minimum threshold when designing their Ph III pivotal trial.

This dose dependent relationship involving CD34+ stem cells has been found before in meta-analyses:

Randomized controlled trials (OTCPK:RCTS) of BMSC-based therapy for STEMI, delivered within nine days of reperfusion, were identified by systematic search. 29 RCTs enrolling 1,830 patients were included. Intracoronary BMSC therapy resulted in an overall improvement in left ventricular ejection fraction (LVEF) of 2.70% [95% confidence interval (NYSE:CI) 1.48-3.92; P<0.001] in the short term and 3.31% (95% CI 1.87-4.75; P<0.001) longer term. Meta-regression suggested a dose-response relationship between quantity of CD34(+) cells delivered and increase in LVEF (P=0.007).

Between their Ph I and recently unblinded Ph II, as well as the above over 1,800 patient meta-analyses, it's clear that a dose-dependent relationship exists between CD34+ stem cells and clinical outcome.

Myocardial Perfusion

One of the co-primary endpoints for the NBS10 Ph II trial was perfusion. However, the full set of primary endpoints listed on clinicaltrials.org for this trial is as follows:

Primary Outcome Measures:

To determine safety and efficacy of intracoronary infusion of NBS10. [Time Frame: primary outcome measured at 6 months] [Designated as safety issue: Yes]
The primary endpoint includes the occurrence of AEs, SAEs and Major Adverse Cardiac Events (OTCPK:MACE) and the assessment of myocardial perfusion measured by quantitative gated SPECT MPI specifically looking at resting total severity score.
It was an interesting idea to use perfusion as a possible surrogate endpoint in this trial, but it was not the only primary ep, nor was it the most important. Making perfusion a co-primary endpoint only made sense from the standpoint that it could be tested for and found very quickly (as is usually the case with surrogate endpoints), thus showing in short order a possibility of treatment effect. MACE (major adverse coronary events), AEs, SAEs, and of course overall survival all take longer to fully register results. So as a surrogate addition to the larger set of more relevant endpoints, it made sense to include it.

What doesn't make sense however is the public reaction to not meeting it. Because what is perhaps most important about this particular endpoint is its unreliability in proving a clinical benefit (extended survival, reduction in MACE, etc.) between groups that have both already had an AMI, which is why it's never used as a primary endpoint in these trials. The purpose of SPECT is to show whether or not you have coronary artery disease (NYSEARCA:CAD) in the first place, and although often effective at that, it still misses a large percentage of the time. It was never meant to test subtle differences between two groups of post-heart attack patients. The detection of the formation of new, very small capillaries is beyond the reach of SPECT imaging, which was used as a measuring tool in this trial.

A single-photon emission computerized tomography (SPECT) scan allows doctors to analyze the function of some internal organs. A SPECT scan is a type of nuclear imaging test, which uses a radioactive substance and a special camera to create 3-D pictures. It can show how blood flows in and out of high volume areas, such as the brain and heart, but it has little capacity for detecting fine details.

Some of the drawbacks of SPECT imaging have been enumerated in studies:

SPECT MPI has multiple limitations, including relatively long acquisition protocols and considerably poorer spatial resolution than other available modalities, limiting detection of subendocardial perfusion defects. Furthermore, the roll-off of tracer uptake at higher myocardial blood flows limits sensitivity in detecting mild-to-moderate stenoses. Additional limitations include motion artifacts related to patient and respiratory motion, scatter and partial volume artifacts in the inferior wall related to gut and biliary activity, and variable attenuation artifacts resulting from breast or subdiaphragmatic attenuation. These artifacts can decrease the diagnostic utility of the perfusion images.

There are apparently better methods for detecting myocardial perfusion, as the above link goes into detail on. PET MPI is one such method, although not yet FDA approved:

PET MPI has been extensively evaluated and has been shown to be both sensitive and specific for the diagnosis of CAD. A recent meta-analyses including 14 studies (840 patients) demonstrated a sensitivity of 0.92 (95% CI, 0.90 to 0.94) and specificity of 0.85 (95% CI, 0.79 to 0.90) for the detection of CAD. The average prevalence of CAD in these data sets was 77%, and the studies included both N-13 ammonia and Rb-82 as the perfusion agents.

In a head-to-head comparison with SPECT, PET showed greater accuracy:

There are multiple studies directly comparing the diagnostic utility of PET as compared with SPECT that have demonstrated similar or superior diagnostic accuracy for PET. A recent article studied 112 patients who underwent Rb-82 PET and 112 patients who underwent SPECT and compared their diagnostic utility, using angiography as the gold standard. The patients were matched for their baseline characteristics, and the cohorts included a subset of patients with low likelihood of coronary disease. On a per-patient basis, PET had a higher diagnostic accuracy (91% versus 76%) and higher specificity (100% versus 66%) for detection of a 50% or greater coronary artery stenosis.

So we see that SPECT was unable to detect 50% or greater stenosis 36% of the time. No doubt in a comparison between two groups of severe AMI patients, the formation of new, tiny capillaries in the one, which would marginally improve blood flow to the area, would likely be obfuscatied by the overall massive blood flow in that region. If SPECT cannot be relied on to find 50% or less coronary artery stenosis, how can it be relied on to find this?

What is of note, however, is that left ventricular ejection fraction (LVEF) was found to increase in the treatment arm. That is a more clear indication of increased cardiac function (and one that regulators accept, unlike perfusion), which SPECT can and has found was improved in the NBS10 arm. This too was dose dependent.

Also, no imaging methodology has the sophistication to tell whether or not CD34+ stem cells had an impact on the repair of the endothelium of the already present blood vessels or the mesothelium of the heart itself. Researchers can only tell by the aftereffects of the therapy, namely, whether or not it had an impact on MACE, overall survival and ejection fraction. These are also the only endpoints acceptable to regulators.

In other words, and contrary to the conclusion jumping of a some, NeoStem's Ph II PreSERVE-AMI study did not fail to meet its primary endpoint - it failed to meet one of four of its co-primary endpoints, which was the least important and only surrogate ep of the group, but only as was capable of being detected with the limited abilities of SPECT imaging (PET may have detected a difference). In the other three of four endpoints, all of which are used and relied upon as the basis for FDA approvals for AMI treatments, NBS10 is showing statistically significant efficacy, which is only likely to increase as the months go by. An early separation of two overall survival curves tends more often than not to grow, especially when early differences are stat sig as they are here.

The recent sell off can be pointed directly towards this misunderstanding of clinical endpoints and how the data should be interpreted. If instead this trial's endpoints were framed more towards what is acceptable to FDA, I think most people would be quite encouraged by these data (as Maxim Group is, for instance). Instead, many are illogically disenchanted. But in fact, the data are encouraging and quite positive, even at this early interim stage (6-12 months for all patients). The only mistake made by management was in deciding to list perfusion as a co-primary rather than secondary endpoint. But it does speak to their high confidence in the therapy.

In another six months and then again in 12 months, data will have matured and will be publicly reported, and in my opinion the gap of separation between groups will have widened. However, there are enough statistically significant data already to begin designing trial protocol for a Ph III study testing NBS10 using all that they have learned from this trial. In particular, they will likely raise the lowest dose threshold to 20mil cells (not difficult as they had obtained doses up to 100mil), and control for time to stent implantation. Though, it is encouraging that the NBS10 group showed such efficacy even at this disadvantage. And there is no doubt FDA will approve the initiation of a Ph III trial on the basis of these data.

These data also increase the likelihood of partnering for the further development of NBS10, and it was probably presented under non-disclosure agreement in the recent financing deal (specific mention of "$5mn upon partnering NBS10"). They may already have negotiations underway. I think BAX may be a likely contender.

Baxter International, Inc. (NYSE:BAX) is the only other company with a like product in late-stage clinical development. Their Ph III trial testing intramyocardial delivery (heart muscle injection) of autologous CD34+ stem cells is set to conclude by mid-2016 (which incidentally is being manufactured by NBS's wholly-owned sub, PCT). Their placebo-controlled 150 patient Ph II data were also encouraging:

CHICAGO and DEERFIELD, Ill., July 7, 2011 - New research published online today in Circulation Research found that injections of adult patients' own CD34+ stem cells reduced reports of angina episodes and improved exercise tolerance time in patients with chronic, severe refractory angina (severe chest discomfort that did not respond to other therapeutic options).

The phase II prospective, double-blind, randomized, controlled clinical trial was conducted at 26 centers in the United States, and is part of a long-term collaboration between researchers at Northwestern University Feinberg School of Medicine and Baxter International Inc. The objective of the trial was to determine whether delivery of autologous (meaning one's own) CD34+ stem cells directly into multiple targeted sites in the heart might reduce the frequency of angina episodes in patients suffering from chronic severe refractory angina, under the hypothesis that CD34+ stem cells may be involved in the creation of new blood vessels and increase tissue perfusion.

At six months after treatment, patients in the low-dose treatment group reported significantly fewer episodes of angina than patients in the control group (6.8 vs. 10.9 episodes per week), and maintained lower episodes at one year after treatment (6.3 vs. 11 episodes per week). Additionally, the low-dose treatment group was able to exercise (on a treadmill) significantly longer at six months after treatment, as compared with those in the control group (139 seconds vs. 69 seconds, on average). Angina episodes and exercise tolerance rates were also improved in the high-dose treated group at six months and at one year post treatment compared to the control group.

They would certainly want in on the development of NeoStem's NBS10, as they have a vested interest in a very similar treatment, which may after all end up inferior to NBS10. This would hedge their bet.

Conclusion

What I am saying that I haven't said before in a former article on NeoStem is that now I know the therapy is effective. Statistically significant differences between the higher dose treatment groups and placebo group in all of the major endpoints respected by regulators in unison, and sequentially with respect to dose, doesn't just happen in a large placebo-controlled double-blind trial. The numerous mentions of statistical significance means it is incredibly unlikely to be the result of anything other than effect of therapy, and the only possible confounder was an advantage given to control.

With in-house manufacturing from an increasingly successful wholly-owned stem cell facility, the imminent initiation of a Ph III trial in metastatic melanoma (whose early data show greater effectiveness than Yervoy), the high potential of forthcoming partnerships, a clean balance sheet and multiple analyst targets of $15-$25 a share, proof they can raise money effectively without dilution by enrolling and completing a large 161 patient double-blind trial, and now data that backs up concept of design for NBS10, making positive data readout from a resultant pivotal Ph III trial testing the same much more likely - I'm at a complete loss as to why investors should not enter a position here, so I continue to recommend it as a strong buy.

http://seekingalpha.com/article/2696855-neostems-nbs10-and-data-that-matters-to-regulators