World first IPS cell therapy for Parkinson’s disease

For the first time, researchers have treated a patient with Parkinson’s disease using induced pluripotential stem (IPS) cell therapy. Everything about this proof of concept case study is out of the ordinary, including the patient: a 69-year old retired physician-turned-entrepreneur who also happened to be the philanthropic funder of the research.

S2N Chief Scientist Prof Michael Valenzuela breaks down the report published in the May 14 edition of the prestigious New England Journal of Medicine.

The Cells

In 2014 fibroblasts from the patient’s skin were genetically modified to upregulate expression of the classic “Yamanaka four” factors, OCT4, Sox2, klf4 and c-myc. The team had earlier reported on a non-integrating virus-free reprogramming method that used episomal vectors to also deliver a microRNA modification cocktail designed to improve IPS cell efficiency and reliability.

Using this, five patient-specific IPS clones were created, and one chosen (C4), because it displayed the lowest number of somatic gene mutations (k = 92) following DNA fingerprinting (whole genome-whole exome sequencing), none of these were related to common cancer mutations. This C4 cell line was frozen as the working cell bank.

For treatment, the C4 line was thawed and entered the researchers’ bespoke in vitro dopaminergic protocol. This involved a complex and multistage chemical process of floor plate induction, neural progenitor induction, midbrain dopaminergic progenitor (mDAP) differentiation, and finally dopaminergic neuronal (mDAN) differentiation. An interesting innovation included a “spot-based” culture method to space out differentiation of cells in clumps at the intersection of a matrix grid to manage the rampant cell death observed during traditional monolayer culture.

The team was also careful to eliminate remaining sub-populations of IPS-derived cells that may have persisted in a primitive pluripotent state, hence posing a major risk of tumor formation. They used quercitin, a natural compound that is a potent inhibitor of the survivin protein (BIRC5 gene) to effectively eliminate 99.99% of such cells: OCT4+ pluripotent cells dropped from an estimated 17 per 10M cells at D28 post differentiation without quercitin to less than 0.001 cells per 10M with quercitin, without an observable impact on dopaminergic yield.

So how pure and functional were the cells at the end of this tortuous process? At day 28, more than 80% of cells expressed markers for mDAPs (FOXA2, LMX1A), 40% for mature neuronal marker MAP2 and less than 15% definitive dopaminergic neuronal marker tyrosine hydroxylase (TH), the key enzyme critically deficient in the Parkinson’s disease (PD) brain needed to synthesize dopamine.

Cells at this point were harvested fresh and transported by helicopter (!) to the hospital facility for treatment. Clearly, cells at this stage of manufacture were an admixture of mDAPs and mDANs, but also likely to be safe given the genomic screening process and chemical elimination of potentially dangerous undifferentiated cells.

The Patient

This 69-year old gentleman had a 10-year history of progressive and idiopathic PD.

Prior to treatment he reported an average of 3 hours of “off time” per day featuring poor motor control and debilitating tremor and posture, and an average daily use of 904mg of levodopa equivalents (total pharmacological dopamine-related drug required for symptom control).

His background medical history was not reported but his social history was notable for being a highly mentally active person involved in the study in major way - first as philanthropic founder, then as first-in-human patient. More about the interesting social and ethical aspects of the study can be found here.

How he was treated

One of the many fascinating aspects of this case study was that it occurred under FDA authority as an Investigational New Drug (IND) under the Individual Patient Expanded Access program.

4 million day 28 dopamine-differentiated C4-IPS cells were transplanted into both the left and right posterior putamen, spaced 6-months apart in two rounds of surgery by FDA requirement. These were done in 2017 using MRI-guided stereotaxis with cells deposited in 3 sub-targets within the putamen.

The patient was in hospital for 24 hours and discharged home the next day.

What they found

First and foremost, at 2-years follow-up there were no adverse event or onset of dyskinesias, so the treatment in this single patient appears to be safe. Second, clinical reports from the patient indicated a small global improvement and halting of long term decline. For example, he reported now only 1 hour of “off time” a day compared to 3 hours pre-treatment and his self-rated quality of life scores improved in a major way.

Clinician rated instruments corroborated that impression, PD motoric scores (MDS-UPDRS III) starting at 38/132 (more is worse) at pre-treatment, varying from 19-35 during follow up and ending at 29 at 2 years. So a modest but I think clinically meaningful improvement.

Biomarker and objective data were, however, not so positive. For example, levodopa equivalents reduced by only 6%, well within the margin of error. 18F-DOPA PET imaging was very unclear. On the left side (treated first), synaptic dopamine noticeably deteriorated at 3-months follow up, and at 24-months ended up between -1.6 to -4.8% worse. On the right side (treated second), final scan results at 18-months ended up anatomically uneven , ranging from 5.4% better in the back half of the putamen (where treatment was delivered) to -4.0% worse in the front half. In other words, dopamine production in 3-of-4 striatal subregions deteriorated.

For the sake of completeness, structural MRI scans also showed clear evidence of gliosis in the white matter tracts lying on top of the striatum, through which the surgical needle passed. This is probably unavoidable given the anatomical target in question and unlikely to be of clinical significance.

Take Aways

This is a fantastic study because it has broken new ground in so many respects, from cell differentiation to neurosurgical delivery. It’s the first IPS cell therapy for PD to be approved by FDA and has laid bare all related cell manufacture and safety implications. It is also important to appreciate that it has been at least 7-8 years in the making if you factor in all the preclinical lab work required to develop their IPS protocols.

Clinical outcomes were modest and of course coloured by the open-label and single-subject nature of the trial. However, the patient’s reports were consistent with clinician-rated tools, but they too were well aware of the intervention and may have been subconsciously biased. Digital tools to objectively capture motoric signs in a high fidelity, continuous and naturalistic way will need to become part of such Phase I open label trials in the future.

Biomarker data were even less convincing. I would go to the extent to say that it is not clear whether functional grafts were established bilaterally given the DOPA activity in the left side deteriorated compared to pre-treatment baseline. Hence, the lack of a strong clinical response, as witnessed by some patients in historical trials of fetal engraftment, may have been due to lack of functional target engagement, i.e., in situ dopaminergic neuronal differentiation throughout the putamen.

One of the reasons for this is that IPS-based cell technology for neuronal cell production remains mercurial. Any given mother stock line will be biased down one germ lineage or another, and it progeny towards one cell fate more another. This particular C4 IPS clone had less somatic mutations than the other 4 clones; what would have happened if one of the others had been chosen on different genomic criteria? And even with a lot innovation around dopaminergic neuronal differentiation, after 28 days less than 15% had the mature TH-positive on-target cell phenotype. Putting all this together, it’s still a bit of a lottery for patients when contemplating IPS cell therapy.

That’s not to say it is impossible. What is needed first and foremost is patient-to-patient reliability and predictability of cell manufacture. A target of >65% TH-positive neurons in the cell therapeutic dose is reasonable and, more importantly, less than 15% coefficient of variance from line-to-line. At this time, the field is not there.

It is also inevitable that any IPS technology is going to require a high load of genomic screening to pass regulatory approval, due to fundamental changes engineered into the cell’s proliferative and transformational biology (by definition). Similarly, ensuring that there are essentially no undifferentiated (OCT4+, Sox2+ etc) cells in the mix is de rigeur.

Creating a multifunctional monster and then trying to tame it for an exquisitely specialist purpose is always going to be hard.

S2N’s technology for isolating a native highly-potent and signal-governable neurogenic cell from the hair follicle, and then enriching and expanding them, is of course, another approach that is GM-free.

S2N is currently undertaking R&D to create a clinical-grade dopaminergic neuron differentiation protocol to help support Parkinson’s researchers in their important work.