Early Clinical Development Strategy: Demonstrating Proof ... · Biomarkers are also used throughout drug development.\爀屲In clinical studies, we use biomarkers to demonstrate the
15
Early Clinical Development Strategy: Demonstrating Proof-of-Concept With Innovative Biomarkers and Focused Patient Populations 1 James W. Kupiec, MD Chief Medical Officer July 2019
Early Clinical Development Strategy: Demonstrating Proof-of-Concept With Innovative Biomarkers and Focused Patient Populations
1
James W. Kupiec, MDChief Medical OfficerJuly 2019
Presenter
Presentation Notes
Hello, this is Dr. James Kupiec, Chief Medical Officer at ProMIS Neurosciences. Thank you for clicking on this presentation.
Early Clinical Development Strategy at ProMIS
Pursue rapid, cost-effective development of our assets that creates an early inflection point in program value
2
Utilize innovative, informative fluid-based biomarkers to rapidly show a disease-modifying treatment effect
Focus on disease populations in each program that have the highest odds of showing a biomarker response in an early, relatively small clinical study
Presenter
Presentation Notes
I’ve just returned from the Alzheimer’s Association International Conference in Los Angeles, and it was certainly very exciting to see and discuss all the new data on fluid-based biomarkers for neurodegenerative disorders and learn how they will certainly impact cinical development strategy. Here’s the agenda for today’s presentation. First, I’ll discuss the impact of these biomarkers on the early clinical development of potential disease modifying therapies. The good news is that a number of these brain-based biomarkers can now be assayed in the blood, and this will contribute to ease of sampling and clinical trial speed. Second, I’ll give three examples to show how selecting an appropriate and focused patient population is critical to maximizing the odds of success in revealing an early assessment of a therapy’s true potential.
What is a Biomarker?
3
A characteristic that is objectively measured and evaluated as an indicator of:1. Normal biologic processes,2. Pathogenic processes, or 3. Pharmacologic responses to a therapeutic intervention NIH Biomarkers Definition Working Group 2001
Biomarkers must undergo a rigorous validation process after identification: Establish relevance to the targeted
population Identify clinical utilityWidespread adoption
Presenter
Presentation Notes
But first a few basic comments about biomarkers. Biomarkers are objective, quantitative measurements that reflect some aspect of human physiology or disease, or even confirm how a disease can respond to a particular therapy. The table on the left shows a number of biomarkers that we are all quite familiar with, be it blood pressure measurements, or blood glucose and cholesterol levels. They have been in use for decades and they therefore have widespread acceptance by physicians, patients and regulatory authorities.
Biomarkers can enhance and accelerate drug development
4
Monitor the safety of a therapy Predict patients who might respond better Potentially enable time and cost savings in trials Determine if a treatment is having the desired
effect on the body
Clinical Biomarkers can:PharmacoDynamic Biomarker• Show a drug hits its target with
an biochemical impact• Demonstrates proof of the drug’s
mechanism of action• Can help determine dose and
dose schedule
Presenter
Presentation Notes
Biomarkers are also used throughout drug development. In clinical studies, we use biomarkers to demonstrate the safety of an investigational compound, to help select ideal patients for a particular study, and we use them to demonstrate that our treatment is having a measurable effect on the pathologic disease process. We call the latter pharmacodynamic biomarkers because they dynamically respond to the pharmacologic action of an investigational drug. The problem, in neurologic disorders, is that most of the really important biomarkers are large proteins. In years past, we’ve been severely hampered by the fact it’s difficult to accurately measure brain proteins that are excreted into the bloodstream. Blood levels are very low and the proteins tend to get degraded very rapidly in the blood. That has been the case until just the last few years. New immunologic or mass spectroscopy techniques now permit measurements at 1/100th or 1/1000th the concentration we typically have been able to measure.
Two examples of a pharmacodynamic marker (NfL) in neurology
5
Neurofilament light in cerebrospinal fluid following monthly administration of natalizumab (300 mg) in 92 patients with multiple sclerosis
Gunnarsson, et al Annals of Neuro 2011Winter B. et al, J Neurol Neurosurg Psychiatry 2019
Tysabri in multiple sclerosis Spinraza in spinal muscular atrophy
Presenter
Presentation Notes
Let me give two examples of a great new pharmacodynamic biomarker called neurofilament light or NfL. This protein stabilizes the shape and function of axons. That’s the part of the neuron that travels far though the brain tissue to connect and communicate with other neurons. When axons and neurons are damaged or destroyed, NfL goes up in both the cerebrospinal fluid and also in the blood. The correlation of NfL levels between CSF and blood is incredibly and surprisingly good. But NfL is not specific for any particular neurodegenerative disease because NfL tends to increase in almost all of them, some higher than others. On the left panel you see that at 6 months after initiation of Tysabri in a group of patients with multiple sclerosis, the NfL levels went down to normal levels. This occurred at the same time the patients improved clinically. The same is true on the right panel for the young patient administered Spinraza for spinal muscular atrophy. The NfL levels tracked the clinical response and plummeted from a very high level after just a couple of doses. Although NfL in the examples above were measured in CSF, we now know from numerous studies that there is equivalency between CSF and blood-based measurements which use the new ultrasensitive assay techniques.
There’s a revolution in available fluid-based biomarkers that reflectneurologic disease pathophysiology
6Courtesy, Prof. H. Zetterberg, U. of Gothenburg
The research community is at the cusp of a new era by having highly sensitive blood-based biomarkers that will enable:(1) the earlier screening for disease, and
(2) the efficient monitoring of disease intervention early in the drug development process
Although NfL appears to be ahead of the pack so to speak in demonstrating its value to monitor disease and drug response, there are other biomarkers that show great promise. In red, I’ve circled a few proteins I’m quite interested in. They all reflect a different process inherent to a particular disease. For example, neurogranin is a protein that contributes to dendritic spine structure and function, and it’s critical for memory consolidation. It is quite elevated in the CSF of patients with AD, but – surprisingly – not other neurodegenerative disorders. There’s a new blood assay for phosphorylated tau that was discussed during a number of scientific sessions at this year’s AAIC. Its strength as a biomarker reflecting neurofibrillary tangle formation or drug response is just beginning to be understood. SNAP25 reflects activity at the terminus of axons, and is clearly modified when diseases attack the synaptic connections of the brain. There are other cells in the brain that are activated in response to a disease, such as microglia and astrocytes, and we can also measure their activity and damage. Biomarkers, both blood and CSF, now allow us to peer inside the brain to assess the damage. In fact, many colleagues in the research community are calling this panoply of various blood and CSF biomarkers a “liquid biopsy”.
Neurofilament light (NfL) is a measure of the rate of neuronal death…..
7
NfL
Bloo
d Le
vel (
pg/m
l)
60
40
20Normal
Pathological
MCI-AD
FTD/ALS
AD
Sources: from AAIC 2018 ABBVIE, Wash U, U Sorbonnne; Rohrer et al, 2016 AAN; Mattson et al, JAMA Neurology 2017
From JAMA Neurology, April 22 2019
“ The findings suggest that plasma NfL can beused as a noninvasive biomarker associated with neurodegeneration in patients …and may be useful to monitor effects in trialsof disease-modifying drugs”
Mattsson, et al, JAMA Neurology
MSA
Presenter
Presentation Notes
Let’s go back to NfL. It’s elevated in patients with AD and correlates with atrophy of brain tissue. NfL levels continue to increase as cognition worsens in AD. It is elevated in multiple system atrophy, an atypical Parkinson-like disorder. We find very high levels in amyotrophic lateral sclerosis, fronto-temporal dementia, various types of stroke, anoxic encephlopathy and active HIV dementia. A very important paper was published 3 months ago evaluating NfL in patients with AD, or MCI due to AD, using thousands of samples from the ADNI study. They clearly showed the value of NfL to reflect the underlying degenerative process and the authors stated this biomarker may be useful to monitor the effects of disease-modifying drugs in patients with Alzheimer’s disease. Well, all of my development friends at other companies that I spoke to during AAIC indicated they were already collecting NfL in their earliest studies. In fact, one key member of the research community called NfL the future “cholesterol of AD”. That may or may not be the case, but, at the very least, it’s an incredible biomarker with huge potential. If we hypothesize that an effective drug should diminish neuronal death and axonal injury in advance of a more typically recognized clinical benefit, then we should be able to show, early on, a reduction in elevated NfL levels.
Early Clinical Development Strategy at ProMIS
Pursue rapid, cost-effective development of our assets that creates an early inflection point in program value
8
Utilize innovative, informative fluid-based biomarkers to rapidly show a disease-modifying treatment effect
Focus on disease populations in each program that have the highest odds of showing a biomarker response in an early, relatively small clinical study
Three examples: different neurodegenerative disorders, each associated with misfolded, aggregated proteins that cause synaptic loss and neuronal death
Presenter
Presentation Notes
Let’s shift gears. Let me give you three examples where being smart and testing an investigational drug in a specific disease population should maximal the odds of demonstrating proof-of-concept early in the development process. Neurodegenerative disorders are caused by misfolded, aggregated proteins and ProMIS has generated highly selective antibody therapies to target these proteins. In each case, abnormal fluid-based biomarkers should demonstrate a correction in response to therapy.
Example #1: Multiple System Atrophy (MSA)
9
Marques, TM, et al. NEUROLOGY 92(13) March 2019
• Aggressive, rare neurodegenerative disease driven by toxic-misfolded, aggregated forms of α-synuclein
• Attacks oligodendroglia and neurons• One of several synucleinopathies (others include PD and LBD)• Symptoms: autonomic dysfunction, parkinsonism and
cerebellar ataxia• NfL markedly elevated
Serum NfL highly discriminates atypical parkinsonism disorders (including MSA) from Parkinson disease or healthy controls
N=29 (22 – MSA, 7 – PSP)
ProMIS has multiple antibodies with highly selective binding to the toxic form of α-synuclein and strong
binding to brain tissue from patients who died of MSA
Multiple System Atrophy
Presenter
Presentation Notes
Example #1 MSA is an atypical parkinson-like disorder that is rare and aggressive. It is caused by misfolded alpha-synuclein, the same abnormality that produces Parkinson’s disease and Lewy Body Dementia. NfL is quite elevated in MSA. Since we now have a blood-based assay, one can measure NfL as frequently as once a month in order to determine how quickly a patient responds to the therapy. With a therapy that is truly disease-modifying, one should observe a normalization of NfL over time. ProMIS currently has multiple antibodies demonstrating great selectivity for the toxic misfolded form of alpha-synuclein, and a study in MSA patients could facilitate proof-of-concept for our antibody as early as phase 1. If the lead compound does not impact NfL or the other biomarkers in such a study, then the feedback loop would be shortened and enable us to test a different antibody targeting a different region on the misfolded protein.
Example #2: AD in Down Syndrome
10
Neurofilament light (NfL) concentration by age and dementia status of individuals with Down syndrome• Caused by a third copy of chromosome 21 (extra copy of the
amyloid precursor protein)• Associated with physical growth delays, moderate intellectual
disability and characteristic facial features• Almost all people with Down syndrome by age 40 have amyloid
brain deposits; the majority will show AD symptoms in their 50’s and 60’s and ultimately develop dementia
• Highly motivated DS community for clinical research; established clinical network of research sites (US and EU).
• NfL is elevated in DS and increases rapidly after age 40• Presymptomatic DS patients with elevated NfL identified and
available for interventional studies
PMN310, the ProMIS lead candidate, has highly selective binding to the toxic form of β-amyloid and strong binding
to brain tissue from patients who died of AD
Presenter
Presentation Notes
Example #2 Patients with Down Syndrome have the highest genetically based incidence of Alzheimer’s disease. They have a third copy of chromosome 21, and that chromosome genetically codes for the amyloid precursor protein. Thus these patients begin to accumulate beta-amyloid in their brains earlier and at a faster rate than patients with sporadic AD, and the large majority of them go on to develop dementia due to Alzheimer’s disease. There are already large cohorts of patients who have been assessed as well as a clinical trial infrastructure in the US and the EU. We plan to work with the Down syndrome research community to assess our lead compound PMN310, and determine whether it has the potential to benefit patients.
Example #3: Amyotrophic Lateral Sclerosis
11
Neurofilament light chain in serum for the diagnosis of amyotrophic lateral sclerosis (ALS)• Aggressive disease that causes the death of neurons controlling
voluntary muscles• Characterized by stiff muscles, muscle twitching, muscle
weakness and atrophy, difficulty with speaking or swallowing• Majority of sporadic ALS patients (up to 97%) have brains
containing TDP-43 protein deposits at autopsy, suggesting its pivotal role in ALS; the current thinking is that TDP-43 pathologic oligomerization and aggregation leads to these deposits
• Highly motivated patients for clinical research• Well established clinical trial networks in the US and EU
ProMIS has multiple antibodies in preclinical development with highly selective binding to the toxic form of TDP-43 and strong binding to brain tissue from patients who died of ALS. ProMIS is in active partnering discussions with large pharma.
Presenter
Presentation Notes
Example #3 Patients with ALS have significant damage to their axons and thus very elevated levels of NfL. This is a horrible, aggressive disease, and the majority of patients have pathology consistent with misfolding and aggregation of a protein called TDP-43. And BTW, this same protein is misfolded and deposited in patients having fronto-temporal dementia, or FTD. ProMIS has a number of highly selective antibodies targeting TDP-43. We know that it strongly binds brain tissue from patients who have died of ALS. So again, by assessing the impact of our antibody therapies on NfL levels in phase 1, we can come to an early decision as to whether to continue to invest in this therapeutic approach or try something different.
Phase 1 trial design concept: in higher dose arms, biomarkers can givea signal suggesting therapeutic benefit early in clinical development
12
0.3mg/kg - HNV
Dose Escalation Design
1 mg/kg – HNV/AD patients
3 mg/kg –AD Patients
10 mg/kg –AD Patients
20 mg/kg –AD Patients
80 mg/kg –AD Patients
3-month placebo-control, then 9-month open-label extension
Growing list of predictive biomarkers under development
to consider
HNV = healthy normal volunteers
40 mg/kg –AD Patients
Dose levels are representative examples, drawn from other neurodegenerative monoclonal antibody programsPossibility to use HNVs in lower dose arms, or as sentinel subjects in each cohort
Presenter
Presentation Notes
OK, I’ve given you three examples of where we can test early in the optimal population. In phase 1, as you can see on this slide, we demonstrate the safety of our therapy at low doses and only escalate up to the next higher dose level in a staggered fashion when we’re satisfied the current dose is safe and well tolerated. We typically continue to escalate until we reach the maximum tolerated dose. We will conduct part of our phase 1 studies in a double-blind fashion, then shift to an open-label extension where each individual knows that they are receiving active drug and not a placebo. By designing the phase 1 study this way, we can also collect blood samples for up to one year in order to assess biomarkers.
Hypothetical example – Phase 1 biomarker readout
13
NfL
100*
* 100 = patient baseline value
Months1260
Placebo/ Large natural history dataset?
3 mg/kg dose antibody
10 mg/kg
20 mg/kg
40 mg/kg
80 mg/kg
Existing historical dataset – use as a “common asset” historical control
A biomarker dose-response would
suggest a significant impact on neuronal
preservation
Presenter
Presentation Notes
There is often existing data for these biomarkers from other sources, sometimes what we call a large natural history database. When one has access to such data, that allows one to compare the change in the biomarker in an interventional study over 12 months and compare that data to what would be expected in that patient population had they not gone on an investigational therapy. A change in the biomarker, particularly a dose-responsive change, would strongly suggest a significant impact on neuronal preservation.
A new development paradigm for neurodegenerative diseases will dramatically improve cost, risk….and success
Preclinical models based on toxic oligomer neurotoxicity
- in vitro, in vivo
Phase 1 studies with biomarker initial POC
Phase 2 studies with biomarker and clinical POC
Dramatically accelerated
feedback loop
Earlier inflection point in
program value
Presenter
Presentation Notes
Drug development for neurodegenerative disorders is going to look a lot different very soon. In the old model, we would frequently move into phase 3 without much phase 2 evidence. A very high-risk decision, and a very expensive one too, The new model of development, as you can see on the right, focuses on early biomarker assessments that will permit a clear Go / No Go decision for subsequent development and investment. All this for about $5 to $10 MM in phase 1. No longer will we need to wait until a futility analysis in phase 3 to make such a decision. That model is clearly no longer sustainable.
Clinical Development Strategy at ProMIS: summary of key points
Pursue rapid, cost-effective development of our assets that creates an early inflection point in program value
15
Utilize innovative, informative fluid-based biomarkers to rapidly show a disease-modifying treatment effect NfL, GFAP, p-181Tau (plasma) Neurogranin, SNAP25, YKL-40 etc (CSF)
Focus on disease populations in each program that have the highest odds of showing a biomarker response in an early, relatively small clinical study Anti-alpha-synuclein Multiple System Atrophy Anti-amyloid beta AD in Down Syndrome Anti-TDP-43 Amyotrophic Lateral Sclerosis
The impressive, recent development of fluid-based biomarkers represents a revolutionary advance in the development of drugs for neurodegenerative disorders and one that will quickly accelerate the assessment of
disease-modifying therapies in early phase clinical development
Presenter
Presentation Notes
Here’s my last slide. I’ve reviewed the shift in development based upon the new availability of fluid biomarkers and their impact on decision making. I hope I’ll have the chance in the future to discuss in more detail some of the other exciting biomarkers listed on this slide. In developing investigational drugs in this field, it is also imperative to select the optimal disease population in order to rapidly test the therapeutic hypothesis. I’ve given the three examples that we at ProMIS are considering as we move our portfolio forward. It’s really not hyperbole to state we’re in the midst of a transformational drug development revolution that wasn’t even imagined 4-5 years ago. These new biomarkers have the tremendous potential to measure drug treatment effect. They will facilitate a faster assessment of potential therapies at much lower costs, and this is something I’m really excited about. Thank you again for joining me today. This is James Kupiec, CMO at ProMIS Neurosciences. I sincerely appreciate your time and attention.
