Amarantus BioScience Holdings (OTCQB:AMBS) stock has been on a wild ride over the past two months. The shares were trading at only 7 cents in May 2014 before news broke that the company would be presenting data on LymPro, the company's Alzheimer's disease (AD) blood test, at two medical conferences in July 2014. The excitement around the potential for the data sent the shares soaring from 7 cents to almost 20 cents, only to see things crash down nearly 50% after the second presentation on July 31st in New York. Below is a chart showing the wild ride in Amarantus shares. We do our best below to explain what the data means and why we believe the market may have misinterpreted the results on the 31st. We remain optimistic on the future development of LymPro.
Quick Background On LymPro
Amarantus is developing LymPro, a blood test for the diagnosis of Alzheimer's disease. We think there is a significant market need for a product like LymPro. The impetus of an AD diagnostic is to distinguish between patients with Alzheimer's and other degenerative dementias such as Parkinson's disease dementia (PDD) or psychosis, mild cognitive impairment, and/or age-related memory loss. A simple and economical test that can accurately and reliably do such would have a number of benefits in the clinical and commercial marketplace, including:
1) Significantly expedite diagnosis, allowing for earlier treatment in the primary practice setting. All current treatments, and most of the emerging ones, can only delay the onset or slow the progression of AD. An overwhelming number of researchers and publications point to early diagnosis as critical to extending a patient's quality of life. Diagnosis at or before the first symptoms could add years to a patient's well-being.
2) Significantly simplify the diagnosis process, allowing for meaningful economic savings to the healthcare system. The current diagnostic paradigm, known as a differential diagnosis algorithm, is a complex diagnosis of elimination. The physician starts with asking the patient or caregiver to provide a medical history and progression of symptoms, and may also make use of psychological studies and tests of the patient's sensation, cognition and motor functions. The physician may order a head CT or positron emission tomography (PET) scan, or magnetic resonance imaging (MRI) to determine if the patient shows signs of atrophy or other changes in the white matter of the brain. They may also order a magnetoencephalograph ((MEG)) to document nonspecific changes in the brain.
These tests are expensive and not always available in rural or suburban areas of the country. PET systems are large and expensive to maintain, and almost never available at primary care practices. Reimbursement remains a challenge for unproven imaging agents and PET scans. Other diagnostic procedures, including protein assay of cerebrospinal fluid ((CSF)) are painful and highly invasive. A simple blood test analyzed by flow cytometry or through enzyme-linked immunosorbent assay (ELISA) would have enormous time and cost savings over the current diagnosis algorithm. It also allows the diagnosis to be performed by potentially thousands of more physicians at the primary level of care.
For example, Eli Lilly's (NYSE:LLY) Amyvid costs roughly $1,600 per dose. Patients then undergo brain imaging through PET scan, which is an additional cost of another roughly $2,500 per incidence. Besides being rather expensive, Amyvid is ineffective if the patient is not experiencing memory impairment due to the buildup of amyloid plaque in the brain. This suggests that early diagnosis in high-risk patients is unlikely with Amyvid. As a result, CMS recently denied reimbursement for Amyvid. In April 2013, Lilly reinforced its commitment to Alzheimer's diagnostics by acquiring a novel Tau tangle diagnostic program from Siemens Medical Solutions USA. The program includes two investigational PET tracers intended to image Tau (or neurofibrillary) tangles in the brain. However, similar to amyloid, there is no definitive proof that Tau tangles mean a patient has Alzheimer's, and by using PET, Lilly is further entrenching itself on an expensive and inefficient platform.
3) Improve outcomes by allowing for a more focused treatment based on the type of cognitive impairment. Accurately distinguishing between a patient with AD, PDD, and MCI should allow the physician or caregiver to tailor the treatment option specifically for the underlying disease. A patient with AD may be started on Aricept, Namenda, or Exelon, whereas a patient with PDD may be on Sinemet, Requip, Clozaril, or Seroquel. A patient with MCI or AAMI may simply be advised to take aspirin, a multivitamin, calcium, or other over-the-counter remedy to improve focus or memory.
4) Distinguish between patients for clinical trials and therapeutic development. The ability of a test to distinguish AD in a mixed MCI or AAMI population or earlier has significant implications in our view. The recent difficulties in proving that amyloid targeted treatments are effective may potentially be more a factor of the late stage of intervention rather than that the approach is flawed. Using a test to identify and then treat Alzheimer's at an early stage may be the key to making these drugs effective. It also assures the developers, of new targeted therapeutics, are focusing on the patients hypothesized to benefit from the proposed mechanism of action.
We point investors to the high profile failures of Pfizer / J&J's bapineuzumab in 2012 or Eli Lilly's solanezumab in 2012 or semagacestat in 2010. Elan's failures with bapineuzumab have been well chronicled following the Phase 2 data in 2008. However, Pfizer and new partner, J&J, pushed forward in Phase 3, thinking they had identified a sub-group of patients without a gene called ApoE4 through post hoc data mining. Some 4,100 patients, 4 years and $500 million later, the Phase 3 trial still failed. Despite not learning much from their failure with semagacestat in 2010, Eli Lilly's solanezumab failed a Phase 3 study in Alzheimer's disease in 2012. According to the Cleveland Clinic, over 99% of Alzheimer's drugs fail clinical trials.
How LymPro Works
LymPro was developed by Dr. Thomas Arendt and Dr. Jens Stieler of the University of Leipzig, Germany. The concept behind the "Lymphocyte Proliferation" test is that cell-cycle dysfunction / dysregulation might be critically involved in the process of neurodegeneration in Alzheimer's disease. Evidence has been provided showing dysfunction of the cell division cycle as a more general cellular phenomenon of the disease. Peripheral blood lymphocytes, stimulated with mitogenic compounds, were less able to express certain early proliferation markers in AD patients when compared to age-matched controls. The results suggest that a systemic failure of cellular proliferation control might be of critical importance for the pathomechanism of AD.
Essentially, LymPro works by identifying immune-based biomarkers in the blood of suspected Alzheimer's patients. Specifically, Amarantus and the University of Leipzig have identified numerous biomarkers of the disease, including CD3, CD4, CD8, CD14, CD19, CD45, and CD69. The diagnostic works by utilizing a statistical model that evaluates the quantitative information of the protein biomarkers and assigns a probability score called a "stimulation index" (SI). SI is defined as the ratio of biomarker expression after mitogenic stimulation to unstimulated control measure by flow cytometer. The SI score indicates whether or not a patient is disease free, has Alzheimer's disease, or is at high risk to develop the disease.
Amarantus has spent considerable time over the past year validating previous work on LymPro, as well as securing additional rights to next-generation versions of the product. There are essentially two key validations the company has been focused on: Analytical Performance and Clinical Performance.
- Analytical Performance: Analytical Performance is the ability to properly measure a marker. As we noted above, LymPro is designed to measure the concentration of immune-based biomarkers expressed in the blood following mitogenic stimulation. For LymPro to have clinical and commercial utility, the company must adequately demonstrate that it can measure the biomarkers it has identified. This was accomplished in October 2013, with the data recently presented at the Alzheimer's Association International Conference (AAIC) held in Copenhagen, Denmark in July 2014. We remind investors that Becton Dickinson assisted Amarantus with this work.
- Clinical Performance: Clinical Performance is the relevance of the marker in a disease state vs. a control. It is one thing to show you can measure a biomarker accurately and consistently, but it is another to prove that the biomarker has relevance with respect to the disease. Amarantus released results in July 2014 at AAIC showing statistically significant differences in various patients with Alzheimer's vs. aged-matched healthy controls across a number of univariate and multivariate models. This is important because the more biomarkers Amarantus can find that are associated with the disease, the more accurate the test will become.
Recently Data Presentations - It's All About Sensitivity
When developing a commercially available diagnostic, accuracy is clearly the most important endpoint. Accuracy is simply how often the test is right. The more accurate the test, the more often physicians use it to assist in the diagnosis of their patients. However, accuracy is a composite endpoint made up of two components: Sensitivity and Specificity. Sensitivity is the "True Positive" rate; or simply the percent of time the diagnostic test calls a positive subject positive. Specificity is the "True Negative" rate; or simply the percent of time the diagnostic test calls a negative subject negative. Error rates increase and accuracy decreases when "sick" patients are called healthy by a test and "healthy" patients are called sick. More on Sensitivity and Specificity can be found on Wikipedia.
Normally, for a commercially available diagnostic there is a balance between Sensitivity and Specificity. It is important to be right on both aspects of the test so that healthy patients are not undergoing unnecessary treatment and exposed to unwarranted cost. Similarly, physicians do not want to send a sick patient home without treatment just because the diagnostic test was wrong. The balance between Sensitivity and Specificity can be seen in the graphic below. Any diagnostic test can achieve 100% Sensitivity simply by calling every patient healthy. Unfortunately, that will dramatically lower Specificity and thus overall accuracy. Similarly, if the diagnostic test calls every patient sick, Specificity will be 100% at the detriment of low Sensitivity.
But Amarantus is not developing LymPro to be a commercially available diagnostic - at least not at this stage in the game. The current path of development for LymPro is through CLIA, or Clinical Laboratory Improvement Amendments. CLIA was passed by Congress in 1988 to establish quality standards for all laboratory testing to ensure the accuracy, reliability and timeliness of patient test results regardless of where the test was performed. CLIA is user fee funded and operates under the Centers for Medicare & Medicaid Services financial management operation. However, the categorization of commercially marketed in vitro diagnostic tests under CLIA is the responsibility of the FDA.
Amarantus is seeking to develop LymPro under the CLIA pathway. The product would be intended for the in vitro qualitative determination of a stimulation index in human blood derived lymphocytes in patients with evidence of cognitive impairment to aid in the diagnosis of dementia of the Alzheimer's type. We suspect the test will not be used as a standalone diagnostic (at least yet), but as an adjunct to clinical assessments and other diagnostic evaluations, potentially including MRI or Amyvid-PET, and that a positive test offers clinical utility as an aid to help identify those patients with cognitive impairment who have dementia associated with Alzheimer's disease for clinical trials or a companion diagnostic product.
Under CLIA, Sensitivity becomes the paramount endpoint. Above we noted the failure of some high profile Alzheimer's drugs over the past few years. In fact, pharmaceutical companies have an abysmal record in developing Alzheimer's drugs over the past few decades - 99.6% failure rate according to the Cleveland Clinic. Why? Many experts believe the high failure rate of Alzheimer's therapeutics is due to two factors. Firstly, pharmaceutical companies are going after the wrong targets, mainly amyloid plaque and Tau; and secondly, pharmaceutical companies are enrolling the wrong patients for their trials. The wrong patients include either Alzheimer's patients that are too far gone (i.e. advanced disease) or patients that do not have Alzheimer's at all. Instead, these patients may have other degenerative dementias such as Parkinson's disease dementia (PDD) or psychosis (PDP), some form of mild cognitive impairment , and/or age-related memory loss. As such, going forward, it is important that big pharma find "True Positive" patients for their programs and avoid "False Positive" enrollment.
Results from AAIC are presented below. The data shows the Clinical Performance of the test with roughly 81% accuracy. Investors may have expected higher results, as previous peer-reviewed publications on LymPro (Stieler, et al, 2001) pegged the accuracy rate around 91%.
However, investors should be aware of a few things with respect to the data presented at AAIC. Firstly, it was preliminary data and the trial was not designed for the multivariable analysis, which Amarantus is moving forward with now. Secondly, results were with older version (V1) of the test and not the new enhanced version (V2). Amarantus presented data on V2 at AAIC, concluding increasing concentration of mitogenic stimulants and longer incubation times significantly increase the cell cycle activation of markers. For example, increased CD69 expression appears to increase univariate differentiation. As such, the company believes that multiple parameters can be tuned to optimize assay performance, signal to noise, and maximize differentiation between Alzheimer's disease and other chronic progressive dementia that confound diagnosis.
On July 31, 2014, Amarantus and Brewer Sports International hosted an Alzheimer's-focused summit as part of the C4CT Concussion Awareness Summit held at the United Nations in New York City. As part of the conference, Amarantus presented interim findings from a clinical trial studying two versions of LymPro, an older original version (V1) licensed from the Provista Life Science and a newer improved version (V2). The company is seeking to enroll 72 patients (although we suspect the number may be closer to 100 once completed) and "bridge" the V1 and V2 data. What was presented at AAIC and C4CT was data from V1 only. We believe data from V2 is coming in September 2014. The data below show statistically significant separation on two new biomarkers, CD19 and CD14, with an overall accuracy rate of around 80%.
Below are combined results from the two above biomarkers, CD19 and CD14, presented at C4CT. Amarantus believes that cutpoints in the data can be tuned to enrich a desired population. Meaning, Sensitivity can be tuned up to greater than 80% by sacrificing some Specificity.
From a commercial product standpoint, it is not desirable to manipulate the test to increase one aspect of the analysis because overall accuracy is the balance of both Sensitivity and Specificity. But for enrollment in a clinical trial, when a big pharmaceutical company is clearly more interested in finding "True Positive" patients than worried about missing a "True Negative," fine tuning the analysis to enhance enrollment is a powerful tool under CLIA.
Why Big Pharma Will Want LymPro
A rapid and economical test like LymPro could be used to improve Alzheimer's patient enrollment for future clinical trials. Right now, most pharmaceutical companies are using things like PET and MRIs or MEGs to select for enrollment of a clinical program. These tests have high False Positive rates and cost upwards of $5,000 per patient. To enroll 3,000 patients in an Alzheimer's Phase 3 program, a big pharmaceutical company may need to screen twice that amount. That's 6,000 patients times $5,000 per PET or MEG test, or $30 million in just enrollment costs! Under CLIA, we suspect LymPro will cost around $1,000 per test. If Amarantus can fine-tune the Sensitivity to 90% by adding additional biomarkers to the analysis - remember they got 80% with CD14 and CD19 and previously got 90% with CD69 alone - without significantly sacrificing Specificity, then to enroll 3,000 patients a big pharmaceutical company may only need to screen 3,500 subjects.
Or more likely, the big pharmaceutical company may still screen 6,000 patients, but instead of $5,000 a pop with Amyvid they may screen at $1,000 per pop with LymPro, eliminating 2,500 where LymPro's stimulation index predicts "Not Alzheimer's." The remaining 3,500 may then go onto have PET or an MRI. Under this scenario, using LymPro just saved that big pharmaceutical company $10 million. It is for this reason that Amarantus CEO, Gerald Commission, believes the LymPro opportunity under CLIA is several hundred million. The other potential advantage of LymPro is the test's ability to find early-stage AD patients that have yet to develop meaningful amyloid plaque in the brain. Amyvid works by identifying plaque build-up. Early-stage AD patients may not have built-up enough amyloid plaque to be viewable on PET. This is where LymPro can really differentiate itself.
Dr. Russell Katz, MD, Director of the Division of Neurology Products in the FDA's Center for Drug Evaluation and Research (CDER) agrees with the logic. In 2013 he stated:
The scientific community and the FDA believe that it is critical to identify and study patients with very early Alzheimer's disease before there is too much irreversible injury to the brain. It is in this population that most researchers believe that new drugs have the best chance of providing meaningful benefit to patients." We believe this reinforces the need and desire for early-stage detection, and the potential multi-hundred million dollar opportunity that LymPro represents.
What's Next For Amarantus And LymPro?
More data is coming! Presented at C4CT on July 31, 2014 was interim data from 44 patients using the old version (V1) of the test. Still to come is full data on all 72 patients on both version 1 and version 2. We suspect that the overall number of patients may be closer to 90 or 100 simply because of the first 44 patients enrolled, 34 were healthy controls and only 10 were Alzheimer's. As such, we think management will expand the study slightly just to make sure the randomization going forward is more 1:1. Nevertheless, we expect to see full data on all patients on both V1 and V2 in September 2014. At that time, we will see if the company's hypothesis that increasing the test to a multivariable approach leads to improved predictive outcomes. We will also see if the increased mitogenic stimulation time and concentration implemented in V2 improves the results compared to V1. Remember, what we saw in AAIC and C4CT in July 2014 was under the old approach with the old version. Amarantus has spent the past year developing an improved version (V2) and approach that we will finally see data on in the coming few weeks. This will be a major catalyst for the shares.
Expanded Role With Dr. Arendt / University of Leipzig
On August 1, 2014, Amarantus announced it had entered into an exclusive option to license the intellectual property surrounding the therapeutic concepts of Dr. Thomas Arendt from the University of Leipzig. Amarantus will now have 12 months to analyze preclinical animal data from Dr. Arendt's lab and potentially negotiate a definitive licensing agreement. Work out of Dr. Arendt's lab pioneers the research effort into Cell Cycle Dysregulation in Alzheimer's disease.
Dr. Arendt has filed international patent applications on the concepts of modulating the expression of P16 in the brains of patients of Alzheimer's disease using gene therapy. P16 is a cell cycle protein that assists in maintaining and nurturing synaptic connections that is hypothesized to be at the root of Alzheimer's pathology. Specifically, Dr. Arendt found that P16 is down-regulated in patients with Alzheimer's disease. As such, a potential new therapeutic approach to treating Alzheimer's disease may exist by creating a gene therapy that up-regulates P16. This is the work being conducted in animal studies by Dr. Arendt's lab at the University of Leipzig. Amarantus believes this option agreement opens the possibility of a targeted therapeutic on those Alzheimer's patients that have tested positive for LymPro.
Amarantus stock was hit hard after the presentation at C4CT. We believe investor expectations on the data may have gotten out of line with what was actually presented. The data the market is expecting is not coming until September 2014. At this time, we have no reason to believe the results will be any worse than those presented at AAIC or C4CT. In fact, when the analysis of blood samples using version 2 of the test is conducted, we expect the results will be superior to what was seen in July. We also believe investors need to be aware of the development pathway for LymPro. The current goal is on obtaining CLIA approval before the end of the year. The focus under CLIA is on assisting larger pharmaceutical companies in enrolling "True AD" patients in clinical trials. This is different from what the ultimate commercial application of LymPro will be, but that is still a few years away. The CLIA opportunity is only months away. And based on recent comments by the CEO, we expect the company to provide revenue guidance for LymPro under CLIA when they hold their second quarter earnings call in a few weeks.
Given the 45% haircut in Amarantus stock on July 31, 2014, we believe today is an attractive entry opportunity. Nothing has materially changed with respect to the CLIA opportunity ahead with LymPro. Development programs with MANF and eltoprazine remain on track, and as of August 1, 2014 we estimate the company has approximately $3.5 million in cash on hand, or enough to fund operations into 2015 without significant additional dilution under the Lincoln Park equity agreement.
Disclosure: The author has no positions in any stocks mentioned, and no plans to initiate any positions within the next 72 hours. The author wrote this article themselves, and it expresses their own opinions. The author is not receiving compensation for it. The author has no business relationship with any company whose stock is mentioned in this article.
Editor's Note: This article covers one or more stocks trading at less than $1 per share and/or with less than a $100 million market cap. Please be aware of the risks associated with these stocks.