Last month, medical and operational experts from Medpace’s neuroscience team presented a webinar on the changes in CNS drug development over the last few years. The webinar, which can be viewed in its entirety here, specifically focused on how our deepening understanding of the human genome and pathophysiology has led to fragmented classifications of neurological or psychiatric disorders based on identified genetic markers. Replacing larger classifications of common disorders with a more precisely defined spectrum of individual rare and ultra-rare diseases is changing the face of clinical research and development.
Over the next several weeks, we will build off each topic that was addressed during the presentation. In Part 1 of a 3 part blog series, we will explore how the heritability of neurological diseases and psychiatric diseases is changing the clinical development landscape. From there, we will take a look at new genetic findings in a number of different diseases and how these findings are making us rethink our broad view of how we ought to approach these diseases. Finally, we will explore the history of treatment of rare diseases with CNS manifestations, touching on both successes and failures.
From Family History to Genetic Subtypes
In CNS clinical practice, we have always known that there is a familial inheritance of neurological diseases and studies of twins confirm this belief. Familial cases of CNS diseases have been thought of as being relatively rare and somewhat different than the sporadic cases that we usually see, but they’ve been studied intensely for hope that they could yield some insight into the pathology of the more common sporadic diseases.
Interestingly, while we’ve understood the genetic underpinnings of these genetic diseases, we haven’t always gotten the kind of insight into pathophysiology that we might like. In other words, discovery of the gene mutation doesn’t always explain the neurological disease. Here are a few examples:
- Wilson’s disease – single gene mutation – ATP7B. Wilson’s disease was known to be a disease of copper metabolism, and the gene codes for a copper transport protein.
- Duchenne Muscular Dystrophy – single gene mutation – DMD. The gene is a gene expressed in muscle, but it’s called dystrophin because we don’t really understand exactly how these mutations in dystrophin cause the degeneration and the distribution of weakness in Duchenne’s.
- Spinal muscular atrophy – multiple gene mutations – SMN1, UBA, DNC1H1, VAPB. The combined gene mutations lead to motor neuron loss, but the exact way in which they lead to cellular death selectively in motor neurons isn’t really understood.
- Smith-Magenis Syndrome – Chromosome 17 deletion, ??RAI1. In this development disorder, the actual location and identity of the gene in that deletion is not really understood or how it leads to the disorder.
- Huntington’s disease and Spinocerebellar Ataxia are examples of CAG repeat diseases. Huntington’s disease, which was one of the first neurodegenerative diseases in which we understood the genetic underpinnings, we still—decades later—don’t understand fully how the CAG repeats which leads to the neurodegeneration that occurs in adult life. Similarly, with Spinocerebellar Ataxia, multiple genes are described which can lead to different syndromes, but all with the same neurologic deficits caused by the same kind of CAG repeat.
From Alzheimer’s disease to Amyloid Disease
One of the best illustrations of this transformation from broad categorizations to more distinct subcategories is Alzheimer’s disease. An Alzheimer’s disease diagnosis is generally made clinically and not until late stages. A clinical diagnosis can’t be made in the prodromal early stages or the “preclinical stage”—the stage of Alzheimer’s disease before any clinical signs or symptoms are apparent. Since we have CSF biomarkers and PET imaging, we’ve started redefining the broad category of Alzheimer’s disease to be amyloid disease where patients are accumulating amyloid in their brains. In addition, risk factors like APO E4, APP and Presenilin in the amyloid pathway allow us to identify individuals who have no sign of disease, but when scanned and looked at for biomarkers, are very likely to have the beginnings of amyloid disease. This early detection allows them to be treated beforehand with primary prevention of amyloid accumulation.
The Usual Approach to Common Diseases
To understand the underpinnings of common diseases, there are generally four areas of concentration.
- Start with the syndrome: Explore the behavior, signs and symptoms, and where it’s localized in the brain
- Look at the functional and anatomic substrate of the disease: What parts of the brain, what types of cells are affected, what’s the actual pathology
- Analyze the level of pathogenesis: What are the metabolic pathways, what’s causing cell death or dysfunction, the genetic risk factors, and the environmental risk factors that lead a patient in neurodegenerative diseases, generally late in life, to manifest the disease
- Seek a final common pathway: Such as trying to either reverse the pathology that’s occurring in the brain or to prevent the pathology from expressing itself or, for most of the treatments that we have for neurologic diseases, actually just mitigating the symptoms by improving the function of the systems that have been affected by the disease
Expanding Genetics and the Impact on Clinical Research
In the last couple of years our understanding of the genetic underpinnings of neurodegenerative diseases has changed dramatically. This is now starting to impact how we are approaching doing clinical trials in these diseases. Let’s take a look at new genetic findings in a number of different diseases and how these findings are making us rethink our broad view of how we ought to approach these diseases.
ALS and Frontotemporal Dementia (FTD)
ALS and Frontotemporal Dementia (FTD) present an excellent example. These diseases are considered different but if you look back in the literature, you will see that some patients with ALS had additional features, particularly cognitive decline. The TDP-43 gene, a highly conserved gene, was first associated with tau pathology in some FTD patients. When the gene was looked at in ALS patients, it was actually found in some familial ALS families. We now understand that there are TDP-43 neurodegenerative syndromes across a spectrum from ALS to frontotemporal dementia. The C9ORF72 triplet repeat similarly can cause ALS, FTD or combinations of both. And of course, ALS has its own collection of mutations that have been discovered; SOD1 being perhaps the first in a familial ALS cohort causing ALS. So, we end up with a picture where there are many genes that can produce either in familial or in sporadic disease, all leading to the same syndrome or set of syndromes.
Similarly, Parkinson’s disease has seen an explosion in the number of genes implicated – first from familial Parkinson’s disease and then fruitful studies of genome-wide sequencing in sporadic patient populations. It’s interesting that it is a collection of variant risk factors, but all of these risk factors fall into categories in which investigators have been investigating for some time. We know that there is a synuclein build-up so there are abnormalities in some patients in the α-synuclein gene. Perhaps a surprise that came out of epidemiologic work was that the gene responsible for Gaucher’s disease is also mutated in some patients with Parkinson’s disease. The connection was made because patients’ with Gaucher’s families were found to have a higher incidence than expected of Parkinson’s disease. In addition, the PARK genes (PARK2, PARK6, and PARK7) are linked to mitochondrial function. So we see multiple genes, multiple pathways and multiple ways to cause the same clinical syndrome of Parkinson’s disease. Additionally, Lewy body accumulation can cause dementia. So, we come in from that Parkinson’s disease thinking about are these patients’ risk factors that all lead to the same kind of process, or they actually find that there’s a genetic defect that leads to an actual genetic disease of late life.
An Evolution to Precision Medicine
These kind of findings are making us rethink our broad view of how we ought to approach these diseases as we understand genetic underpinnings. Perhaps they really are not single syndromes to be approached by a common final pathway, but actually treated as separate diseases with different approaches depending on what’s causing the disease in an individual patient.
This has been called the precision medicine approach.
- Start with the syndrome: Explore the behavior, signs and symptoms, and where it’s localized in the brain
- Subgroup the patients based on their genetic diagnosis: What metabolic pathways are involved, what cells are being killed or not functioning well and, possibly other genetic factors that are either worsening, speeding or protecting some patients from the disease. Environmental risk factors including aging are also likely interacting with the genetic background.
- Choose a targeted therapy based on the subgroup to prevent that particular pathology, reverse it, or mitigate it based specifically on what pathology was driving the disease in a particular patient population.
A Look at Rare and Orphan Diseases
In rare and orphan diseases, this has become a familiar approach because those diseases generally start with an understanding that there is a genetic mutation driving the disease which is the obvious target for treatment. Rare disease drug development has been driven in part by the regulatory environment which has encouraged the approval of the development of these very small patient populations. The two areas with the most activity include oncology—primarily because these are somatic mutations that are identified and then targeted—and the broader category of genetic disorders (many with CNS manifestations).
There have been great successes in orphan and rare diseases. Forty-one percent of the 2016 FDA drug approvals were to treat rare diseases. One of the earliest and greatest successes in treating rare diseases was enzyme replacement therapy for inborn errors of metabolism due to genetic defects. Cerezyme® was approved in 1994 for Gaucher’s Type I, non-neuropathic disease. ELAPRASE® is approved for Hunter syndrome (MPS II), but only helps patients without CNS involvement, which is the minority of patients with MPS II. Overall, there have been nine enzyme replacement therapies approved. They’re generally given intravenously requiring weekly or monthly infusions and cost upwards of $100,000 per year for the rest of the patient’s life. In all of the syndromes in which the replacement therapy is possible, when administered intravenously, it generally has little or no impact on CNS or even peripheral nerve function.
Psychiatric Diseases: More Difficult – Slower Progress
In psychiatric diseases, we’ve seen slower progress. One exception is autism which is driven by the genetic abnormalities of Fragile X syndrome. But the evidence for heritability in psychiatric disease is strong—schizophrenia and bipolar disorders are clearly being inherited. The problem has been that the genome-wide association studies (GWAS) and epigenome-wide association studies (EWAS) have generally been inconsistent. Loci are found, often just SNPs, and then they don’t hold up in other populations and have not led to generally targeted therapies. It is not clear why, but we don’t have the same kind of neuropathology that we do in neurodegenerative diseases so it could be that our syndromes are too broadly defined. Depression may just be too many different drivers from a genetic background in order to pull it out of these kind of sites. The other possibility is that some of our diseases are too narrowly defined. One recent GWA study had a much higher signal when more broadly patients were considered in relatives of schizophrenia who also had things like bipolar disease or other psychotic disorders. It may just be that the number of genes affecting behavior in the brain is too large, or the impact of the environment or resiliency is too large in psychiatric disease and this would be a difficult thing for us to actually guide down to targeted therapies.
A New Wave of Treatments – A New Wave of Challenges
Based on this landscape, drug development is undergoing a sea of change. The final common pathway approach which yielded great drugs for many of the diseases in past years, has not shown a lot of success with new targets. Biotechs in particular, but large pharma as well, are beginning to address these kind of small populations based on genetics. We are seeing a variety of approaches including:
- more targeted CNS replacement of missing enzyme function
- suppression of CAG repeats, including allele specific approaches which is subpopulation of the subpopulation in that disease
- intrathecal RNAi
- small molecules targeted for patients with a mutation, for example modulators of GBA in patients who have GBA mutations.
The challenges mirror those in rare and orphan diseases including finding and identifying patients, enrollment and retention, as these will typically be very small trials spread geographically. Genetic testing in diseases where these kind of risk factors are being developed has not generally been done because there is not a lot of genetic counseling and we don’t really understand what risk goes along with each one of these mutations. Of course, many mutations are found in the same gene so you generally need to sequence the gene and may come up with new mutations all the time. In addition, intervening in the CNS has always been a challenge for CNS drug delivery. But now we’re seeing with these targeted therapies some very novel approaches to delivery of treatments that are demonstrating targeted engagement. One of the things that we’ve learned from orphan and rare diseases is that often when you target the actual cause of disease, the effect size in the trial is larger and smaller trials are possible. We don’t know yet whether or not that’s going to pan out but we can be hopeful.
In Part 2, we will share some lessons learned in identifying, enrolling and retaining rare disease populations in clinical trials.