Neuralstem: Initial Thoughts on Just Released Encouraging Phase 2 Data on the Use of NSI-566 Stem Cells to Treat ALS. Neuralstem Will Proceed to a Potentially Registrational Trial in 2016 (CUR, Buy, $1.50)
Introduction
Neuralstem released this morning (September 29) new information from principal investigator Eva Feldman on phase 2 data from the use of NSI-566 stem cells in the treatment of ALS. She also discussed combined results with data from the phase 1 trial. I have included verbatim the information in the press release that discussed the trial results. I then give my initial interpretation of the results. At this point, the information is very topline and I look forward to seeing a more extensive release of data, which I am sure will be forthcoming.
Key Verbatim Text from the Press Release
GERMANTOWN, Md., Sept. 29, 2015 /PRNewswire/ -- Neuralstem, Inc. (Nasdaq: CUR), a biopharmaceutical company using neural stem cell technology to develop small molecule and cell therapy treatments for central nervous system diseases, announced that nine-month Phase II and combined Phase I and Phase II data on the NSI-566 trial in amyotrophic lateral sclerosis (ALS) was presented at the American Neurological Association Annual Meeting by principal investigator, Eva Feldman, MD, PhD, Director of the A. Alfred Taubman Medical Research Institute and Director of Research of the ALS Clinic at the University of Michigan Health. The data showed that the intraspinal transplantation of the cells was safe and well-tolerated throughout the escalating doses, reaching a maximum tolerated dose of 16 million cells via 20 bilateral injections. There appeared to be no acceleration in disease progression due to the therapeutic intervention.
Researchers calculated a 95% confidence limit around the slopes of decline of ALSFRS-r scores, forced vital capacity (FVC) and grip strength of the ProAct historical database subjects, and evaluated if trial subjects fell within or outside those limits. 73% of Phase II patients, and 79% of combined Phase I and II patients, fell above the upper confidence limit of the ALSFRS-r score. 50% of Phase I and II combined, and 40% of Phase II patients' forced vital capacity percent predicted fell above the upper confidence limit, compared to the ProAct database. ALSFRS-r scores correlated most strongly with FVC preservation, which was the target of the cervical injections. For grip strength control, researchers used the Ceftriaxone (CEF) study database, since grip strength data was not available in the ProAct database. 67% of Phase I and II combined, and 60% of Phase II patients, all at nine months post-intervention, fell above the 95% upper confidence limit.
The most common adverse event (AE) was transient post-operative pain due to surgery. One serious adverse event due to the surgical procedure was observed, but was not attributed to the cells themselves. The patient's motor function was initially weakened and then recovered to the patient's ALS baseline.
"Based on this encouraging safety and clinical effect, we look forward to moving to a registration-directed trial in 2016," said Karl Johe, PhD, Chief Scientific Officer.
Key Points
- The most important point was that the Company is going to begin a registration directed trial in 2016. There was no description of the design of the trial and especially if there is a control group. In previous 10-Q releases they have indicated that there will be a control group. My question is whether this control group will be given sham surgery or will be an observational group? At times in the past, the Company has made statements that lead me to believe that this trial might enroll around 50 patients.
- The FDA usually requires two well controlled trials for approval of a new drug, but in life threatening diseases without effective therapy, it can grant accelerated approval based on one trial. If this new trial duplicates the results of the phase 1 and phase 2 trial, the NSI-566 stem cells could be approved for ALS based on just one trial.
- The release said “There appeared to be no acceleration in disease progression due to the therapeutic intervention.” This language is important. There have been a number of clinical trials in ALS (all of which have failed) and in each case the primary endpoint was to show the rate of decline of patients with ALS. Stabilization of the disease is a much higher level of therapeutic achievement.
- Neuralstem in the past has released very detailed data on individual patients; the company has been very transparent in this regard. This press release is the first in which they have compared data to historical controls from the extensive ProAct data base (which is based on results in the control groups of trials done with other drugs). Based on this comparison, they compared the slope of the decline in ALSFRS-r, forced vital capacity percentage and grip strength of NSI-566 cells to the ProAct data base.
- Based on the comparison to ProAct data for the combined group of patients treated in phase 1 and phase 2, the results show very promising improvement as the press release detailed. However, it is very important to understand that there have been 27 patients treated overall. In the phase 1 trial, the first three patients were extremely advanced and there was little chance of improvement. Also in the phase 1 trial there were 2 patients with bulbar disease and there is strong conviction among key opinion leaders that bulbar patients won’t respond to NSI-566 therapy. This suggests that perhaps as few as 22 patients (not 27) should be compared against the ProAct data base. If this were done the results would look much better. The data analysis presented is a very conservative intent to treat presentation and probably understates the effect that we might expect in a relatively early stage ALS patient without bulbar disease.
- The most significant omission in this press release was a discussion of the non-responders in the phase 2 trial. See my March 16 report “A Closer Look at Encouraging Phase 2 Results for NSI-566 Neural Stem Cells in Treating ALS for more detail.
- My hypothesis is that some of these non-responders were bulbar patients. If so, the proper comparison with the ProAct data base would involve fewer than 22 patients and again having the effect of making the data look better.
- This is an extensive surgery which has the potential for significantly harming the patient. However, safety does not appear to be an issue. The press release stated “The most common adverse event (AE) was transient post-operative pain due to surgery. One serious adverse event due to the surgical procedure was observed, but was not attributed to the cells themselves. The patient's motor function was initially weakened and then recovered to the patient's ALS baseline.”
Investment Thesis
Neuralstem will be taking NSI-566 into a trial that will be potentially registrational in 2016. The data discussed in this press release indicated that the NSI-566 stem cells had a significant therapeutic effect when data for all patients treated in the phase 1 and 2 trials is combined. However, some of the 27 patients treated in the phase 1 and 2 trials (certainly 5 and perhaps more) would not be expected to respond because of the advanced stage of their disease or because they were bulbar patients. In a healthier group of early stage, non-bulbar patients that will likely be the subject of the upcoming trial, there is reason to think that results will be even better.
I think that this validates my investment thesis that the NSI-566 stem cells are a potentially paradigm changing treatment for ALS that in a best case scenario could be approved in 2018. The one outstanding question that I have is whether some of the non-responders in the phase 2 trial were bulbar patients and if not, what are the possible reasons for their rapid decline?
I think that the stock should respond strongly to this positive data, but there is a caveat. After the topline phase 2 data was released in March 2015, the stock was the target of a coordinated short selling attack by a cartel of hedge funds. Using illegal naked shorting conducted in dark pools, this cartel can have an enormous impact on stock prices of smaller (and larger) companies. This is not unique to Neuralstem; it is a pervasive criminal activity affecting innumerable companies. A trademark of the cartel is to turn good news into bad news by unleashing some bloggers to question and attack the data and then unleashing a new short selling attack. Jim Cramer of The Street.com described this tactic in his famous interview describing how he and other hedge fund managers manipulate stocks. You may want to watch his blog which is called “Jim Cramer Explains How The Stock Market Is Manipulated.” If the cartel holds true to form, they will try to portray this news as indeterminate or bad and try to drive the stock price down with a coordinated short selling attack to make this appear as bad or indeterminate news.
Tagged as CUR, Neuralstem, Inc., NSI-566 stem cells, Phase 2 trial of NSI-566 neural stem celss + Categorized as Company Reports
Larry,
Thanks for the article. It was actually a total of 30 patients treated in the combined trials, 15 per trial. That being said, your point holds true…
Phase I:
Patients 1-6 = nonambulatory, and thus not applicable to future trials
Patients 13-15 = bulbar, and thus not applicable to future trials.
So, we can actually make the case only 6/15 patients from of the Phase 1 trial are representative of the treatment population. Of which, 5/6 show a near zero decline over +800 days.
In regards to the p2 results released yesterday… not much new data. That being said, patient #315 now appears to be a “responder” with ~.80 decline in ALSFRS/month since surgery, and ~.25 decline in ALSFRS/month from day ~75 through the most recent data of day ~315.
In addition, it appears 4/5 with starting ALSFRS scores > 40 are “responders” with near zero decline, and 6/9 receiving +6m cells are also “responders”.
It will be very interesting to see the published data and demographics of the p2 patients…
Hi Larry,
Thanks for this interesting report.
Looking at the graph from the presentation, it seems that sadly patients 305 & 310 did not make it about 100 days post surgery. do we know the reasons for that?
about 2 weeks ago, Eva Feldman said in a lecture that there was one center in which they had 1 (or more) bulbar patients. is there a way to diagnose a patient as bulbar – non bulbar in an early stage prior to the surgery?
the thing expected in the new data was the separation between responders and non-responders they used in the first report. do they still think there are identifiable subsets of patients that will/ or will not be affected by the treatment?
Thanks a lot,
Tal
This is the worst stock and management I have ever seen. Since the “good” results, we have dropped another 20% on our way to zero. Management is oblivious & criminally indifferent to shareholders. Worse than useless. Instead of buying shares, they sell them. Instead of issuing a “Breakthrough in ALS” PR they issue the most timid tripe that is the minimum legally required.
I can’t believe I fell for this. Sell now for a big loss or go down with the ship?
When you are being attacked you fight back. Neuralstem management bends over & passes the lube with a shy smile
Dr. Feldman previously stated for certainty at least 1 patient in the p2 trial was bulbar.
With that in mind, Ambulatory non bulbar “Responders” appears to be 8/14 for p2, 5/6 for p1, or 13/20 for combined results (65%).
xforum:
Based on eyeballing the graph it seems to me that patient 315 went from 44 to 35 ALSFRS or almost exactly the 1 point/month historical norm Larry has discussed in the past. So, I think it’s hard to identify that patient as a responder. Additionally, and again this is from eyeballing the graph of the 15 Phase 2 patients it seems that the average rate of decline was almost exactly 1 point per month. While 5 patients (not counting the two apparent deaths) declined roughly at twice that rate, there were four that were stable (averaging no decline at all).
In Larry’s March 16th report he reported that Dr. Feldman had said there would be no bulbar patients in the Phase 2 study: “The results were sufficiently disappointing that Eva Feldman said that bulbar patients would be excluded from Phase 2.” Given that she apparently didn’t make any reference to bulbar patients in her speech the other day, it seems safe to state that there were no bulbar patients in the phase 2.
Looking at the data by cohort, I don’t see any obvious patterns. Very roughly speaking each cohort had patients who responded very well and patients who responded very poorly. The four patients that responded the best came from four different cohorts with the 5th and 6th best responders coming from the remaining cohort. Similarly, the 7 patients who experienced the biggest declines (the 5 with the largest numerical declines and the 2 presumed deceased patients) were spread amongst all five cohorts. I’ve spent several hours charting and compiling these numbers and looking at them and honestly, can’t see any pattern.
Another way I looked at the data was by starting ALSFRS scores. Possibly some hope here but the data is very rough. Yes, all four of the (presumed) surviving patients who had 40+ ALSFRS scores at the time of surgery responded wonderfully (all are basically stable). The 5th patient, #305 started with a 44 and declined rapidly to 30 at the time of (presumed) death.
That sounds good, but there were 5 patients with starting ALSFRS scores in the range of 36-39. One of these patients only declined 5 points, but the other four patients were those whose scores declined the most (-23, -19,-17, -17). These were patients #313, 302, 301 and #306.
Continuing on, the one patient with a starting ALSFRS score in the 32-35 range declined 4 points. The patients starting in the 28-31 range (#304 +2, #308 -15, #310 -9 before his/her presumed death, and #312 -6).
So, if starting ALSFRS score makes a difference you might expect it to be more linear.
A couple of other thing I’d like to note. The Ted Harrada cohort in Phase I received 10 lumbar and five cervical injections of 100,000 cells each. Their closest counterpart in the Phase 2 trial, the 5th cohort of patients 313,314,and 315 received 20 lumbar and 20 cervical injections of 400,000 cells each or 10-11x the number of cells in total.
Finally, a reminder that those last 3 patients were the only patients in the Phase 2 to receive lumbar injections. In his March report, Larry pointed out that patients 7,8,9 in the Phase 1 study as well as the Harada group 10,11,12 generally performed well. Patients 7-9 received 5 x 100,000 cell lumbar injections.
I wonder if more cells is not necessarily better and I wonder if lumbar injections are more efficacious for some reason. As I compare the two studies I think part of Neuralstem’s mistake here was they moved on to Phase 2 too early. While it focused on attempting to find the optimum dose it simultaneously explored new ground by focusing on cervical patients. Because they explored 2 variables it seems that in retrospect they over-reached. From my untrained eye it looks like the Phase 2 results are basically noise and there is little indication of efficacy.
Call this bad luck or just plain poor study design, but to me, it looks like Neuralstem just blew itself up. The stock price is going to tank, and unlike Larry, I think rightfully so. The next round of financing will effectively wipe out investors who bought shares at north of a $250m valuation. Our only hope is strong data from the small cell molecule. However, given the mushiness of the data collected in mental health assessments, CUR’s limited financial position, and their poor study design capabilities (factor 3 might be driven by factor 2) I have very limited hopes for Neuralstem.
Let me preface my response by stating a few initial things. #1) There is NO existing treatment for ALS, it’s a death sentence. #2) Based on Proact analysis, we can conclude typical decline is 1.2/month, but with huge stan dev. #3) The phase 1 results demonstrate 5/6 of the target population (ambulatory, non bulbar) show a near zero decline over +800 days. Please remember, the p1 trial consisted of 15 patients, 6 of which were non ambulatory and 3 bulbar. This skews the combined data drastically and until we know the complete patient demographics of p2, we are in the dark.
Now regarding your comments:
– Patient #315, I have 44-36 over ~315 days, this works out to be ~.8/month… but to each their own. More importantly from days ~75 through ~315 the decline was ~.25/month. Per the PR, one patient showed adverse effects due to surgery, but then recovered and improved. #315 fits the description (Kreg Palko).
– Dr. Feldman publically stated in a presentation several weeks back that at least 1 patient was bulbar and should NOT have been included in the trial. Which specific # patient is unknown, but it obviously skews the data negatively.
– The independent researcher who is HIGHLY regarded has stated several times, healthier, early disease progression patients respond significantly better. Based of the ALSFRS scores, 4/5 starting with >40 were “responders” with a near zero slope. It is obviously possible the 1/5 was the bulbar. Regarding your linear comment, NOTHING in ALS is linear.
-Regarding # of cells, I would also disagree: Cohort A 2m cells = 1/3 responder ; B 4m cells = 1/3 responder ; C 6m cells= 2/3 responders ; D 8m cells= 2/3 ; E 16m cells = 2/3 … ie 6m = 66% responders.
-What “mushiness of data collected in mental health assessments” are you referencing? Only topline data has been presented with a peer reviewed pending. Data was strong.
-“poor financial condition” <– they are capitalized until 2017 with a burn rate of only ~4m/qtr.
-"their poor study design" <– Feldman is best in class, wondering how you can justify that statement?
Some of your comments lead me to believe you may have alternative motives…
New Dr. Feldman keynote presentation from 9/17/15 @ Swedish now available: https://videomedia3.swedish.org/Mediasite/Play/bff82373ea0b485e8acf31628684cd2e1d?catalog=ceb2826a-be98-4f19-b4fd-febacd9b6b49
Explains that the patient with adverse event due to surgery (not cells) was actually in cohort C. The weakness lasted ~90 days. <– Per ALSFRS line chart, it's obvious this was patient #308. This explains one of the reasons for a quick decline post surgery in one of the patients. Day 90 – ~275 (most recent data point) shows decline of only ~.65/month. This patient will still be referred to as "non responder" but it's worth noting.
We now know why 2/7 were quick delciners and non responders… 1 bulbar patient which should not have be included, and 1 adverse event which then recovered to slower disease progression.
xforum: Below is my response to your note from yesterday. I don’t have time right now to incorporate the additional info you provided this morning (thanks for posting the link) but hopefully that omission doesn’t detract too much.
I appreciate your measured response to my points and I’ll continue in that spirit. Also, just because I am questioning and skeptical doesn’t mean I’m a troll or a short-seller. In fact, I loaded up on the stock and short-term options anticipating that in her talk Dr. Feldman would explain the obvious concerns that arose from results reported in March. However, instead of explanation, we received what seems to me appears to be a highly contrived statistical analysis. I’ve been a big fan of Larry’s for awhile and am extremely concerned about illegal and unethical short-selling. In fact, even though I’m a relatively small and unsophisticated investor, I also attempt to fight back against the short sellers as I can through blog posts and investing in companies (NWBO is my largest position). However, I also strongly believe that we shouldn’t allow that concern to constrain open and honest dialogue about these companies. If we can’t have that discussion behind Larry’s paywall then there is a real problem, not to mention a substantial diminishment of the value of the site.
That said, 99% of your note addressed my comments in an open and honest way and hopefully, I can respond in what you perceive to be an open and productive way.
Starting at the top of your e-mail:
1. I am aware that there is no treatment for ALS but I strongly believe in openness and honesty. I do not think the disclosures we’ve seen from Neuralstem and Dr. Feldman over the past six months reflect the long-term best interests of current and future ALS patients. I certainly acknowledge that I only know a small part of the total story but we have to work with the information we have and I am selling down my position and am sharing my views in the hope that I can learn from the dialogue and that I can help other investors learn more.
2. I’ll accept the Proact data being 1.2 ALSFRS points a month with a large standard deviation. While that is better than my tabulated 1 point per month, the unspecified large standard deviation makes it unlikely that 73% of the patients (11 of 15) were 2 standard deviations above the mean in the data. If we assume that patients 301, 302, 206, and 313 are the four that didn’t exceed, that still leaves 308 (-15 ALSFRS points or more than -1.5 per month), the two patients who are believed to have died (#305 and #310), and patient #315 whom you describe further down.
3. I agree that the Phase I trial showed great promise and in my note I focused on the same five patients. I then went on to discuss my concerns about how the Phase 2 trial may have been overly ambitious because the investigators didn’t point out any learnings from the Phase 2 and in my view they didn’t generate any compelling data.
Bullet point A. That’s great info and helpful. That provides hope that #315 is a responder. Of course, it also demonstrates that the surgery has risks and presumably raises the bar a bit for the FDA.
Bullet point B. That’s great as well. Both A and B fit in the category of mitigating points that should have been included in the slides.
Bullet point C. I’ll accept that as a correction. Again, if Dr. Feldman had addressed patients #315 and #305 in her presentation then she could have provided some reasonable justification for the view that the treatment can be expected to perform much better on early stage patients. That is, I, and I think her academic audience, would have wanted more information, more insight into her thinking. In my mind it is really stretching credulity that she would have overstepped some boundary if she addressed the issue of the bulbar patient and the one with the surgical complication.
Bullet point D. I think we differ on the mid range patients. Are patients #309 and 312 responders? They declined at roughly .55 and .66 points per month respectively. Accepting the 1.2 number you cite and if we assume the standard deviation is less than .3 points then they would in my eyes be considered responders.
Bullet point E. I’l apologize for using the word mushy to describe the data collected in the mental health assessments. In mental health assessments the assessments are subjective and there is a very large placebo effect. Much better would be some kind of blood test or bio marker. Even the ALSFRS metric is in my opinion, and I think most experts would agree, more robust than subjective psychological assessments.
Bullet Point F. Yes, poor financial condition. The new Phase 2 study with 50 patients is expected to start in early 2016 and presumably release data in July 2017. Given that the Phase 2 officially started in September 2013 and released data in March of 2015 (18 months) and this study is larger and I’m assuming starting in 2016 means January 1st that is a pretty optimistic date. So, even assuming the burn rate doesn’t increase they’ll need to raise money long before results come out. Ideally, they’d want to keep at least 1 years cash in reserve so I think it’s fair to say that they’ll be raising money in the first half of 2016 if not sooner.
Bullet Point G. Presumably Dr. Feldman is a world class researcher given her status as a former president of ANA and as a U of Michigan Professor. However, I don’t know if designing such studies is part of her core expertise and I don’t know what financial constraints she and the company are operating under. The fact that we are having this discussion and how little learning was communicated in the press release and presentation slides (the stock slide would indicate that the professionals who presumably have information about what was said at the presentation weren’t impressed either) is a strong indicator in my view that the study was poorly designed. I’m a layman with no professional medical background whatsoever and it seems very likely to me that the Phase 2 study overreached. The five successful patients in the Phase 1 all had lumbar injections (note the Harada group had cervical injections 18 months after the lumbar injections, long after the fantastic results were achieved). From my limited understanding of these studies, the Phase 2 is usually more about identifying the proper dose and it appears that they aggressively experimented with that variable in the Phase 2. By adding the cervical variable on top of the dosing variable I believe may have been a problem. Clearly, this is not a scientific assessment. I’m just a regular guy trying to think logically with limited information. I hope someone with more knowledge will explain that the lumbar/cervical variable is of extremely minor or no significance and put this concern to rest. Again, I intend no disrespect to Dr. Feldman. I’ll apologize for using the word “poor” and instead use the word “aggressive” to indicate that I think they tried to do to much too quickly.
Great comments.