DeciBio had the pleasure of interviewing Stephane Mouradian, senior vice president of Business Development & Marketing at Personalis. We interviewed Stephane to learn more about the focus of the company moving forward and their highly sensitive NeXT Personal™ assay.
In 2021, Personalis pivoted to focusing on the oncology market, moving away from the population sequencing aspect of the business. With this pivot, what areas are you focusing on?
We are focusing 100% on the oncology market, with both biopharma and diagnostic customers.
What customers are you focusing on?
In diagnostics, we are starting with the large and more advanced cancer centers, such as NCIs that would be interested in using an exome/transcriptome scale platform. This is the place to start. There are more than 70 NCIs in the US. To date, we have announced partnerships with Mayo Clinic and UCSD Moores Cancer Center, and we have additional ones in the works.
In biopharma, we are focusing on biomarker translational and diagnostic leads. Within these areas, we are currently working with most of the top 10 global pharma companies with oncology segments.
Will you continue to work with Biopharma customers, or focus increasingly on clinical customers? Can you tell us how Personalis is thinking about the interplay between their pharmaceutical and clinical customers?
We will maintain a focus on both biopharma and clinical customers. We do anticipate that over time, the clinical market will become a larger market than biopharma. Similar to others in the space, we will always have a dual biopharma and clinical business. As far as the interplay, we think that biopharma customers stand to benefit from the advances we are making on the clinical side.
What is the relative appetite of these customers for larger panels / whole exome sequencing (WES)?
The appetite is strong. Most biopharma customers today, at least on the translational side, work on the exome/transcriptome scale already. What is missing is a track to diagnostics, and the ability to have a continuous path from translational research to diagnostics and companion diagnostics without needing to “bridge” to a new platform. What these customers are really interested in is the ability to do their translational work at the exome scale, and to then be able to carry that over directly to the diagnostic side without having to find and bridge to a new platform. Bridging is expensive and takes time, so we are helping to solve that issue.
On the clinical side, we are finding that customers at advanced cancer centers and NCIs want to work at the whole exome or transcriptome scale to obtain the most insights for patient care. At Mayo Clinic, for instance, the head of the cancer center there wanted to work at this scale. It is really the way to get the most comprehensive view of the biology of cancer.
How do you see Personalis’ position relative to other exome-based platforms?
Our exome-based platform is not a commercial kit. It is proprietary and augmented for many different features. First, we use a proprietary technology called ACE (Accuracy and Content Enhanced) by which we deliver the most uniform coverage across all 20,000 genes. We have probes in difficult to sequence regions of the exome to make sure that we have these genes well-covered and that we are not missing any mutations. That is particularly important in settings such as immunotherapy where you are determining all the new antigens that could potentially affect response to your therapy.
We also augment our ImmunoID NeXT Platform® for features that track mechanisms of action, as well as mechanisms of resistance to cancer. As an example, we boost the HLA gene regions to be able to do HLA typing and determine mutations in HLA at the same time as we run the exome. In immunotherapy, this is especially important to determine some of the tumor escape mechanisms. Publications have shown that up to 30% of patients that do not respond to immunotherapies have mutations in the HLA gene regions, which is why we are tracking that.
We also have additional features on the tumor microenvironment side for T cell and B cell repertoire sequencing, as well as immune cell infiltration from the transcriptome. It is a unique platform where the idea is to be able to put as many tests as possible (genomic testing in addition to HLA typing, T cell / B cell repertoire, immune cell infiltration, etc.) all into one test that we can run from only 5 slides. The reason for doing that is that tissue is always at a premium, and the idea here is to try and provide the most comprehensive view of the patient’s cancer using a minimum amount of sample.
Personalis has launched a tumor-informed monitoring test, NeXT Personal. What do you think the value is of tumor-informed assays vs. tumor-naïve assays?
In the therapeutic space, there is a strong drive towards detecting cancer recurrence in increasingly earlier stages of cancer. The reason for doing that is if you can identify the presence of cancer at an earlier stage, that is when it is the most treatable. This puts a premium on sensitivity and the ability to detect pathological cancer at the lowest level of ctDNA.
Tumor-informed approaches are more sensitive than tumor-naive approaches, and that is why we have chosen to go that route. At the same time, we were not interested in just providing a tumor-informed approach that would be as sensitive as other existing approaches because we believe the existing approaches are still missing patients who have cancer. We see a fair number of false negatives in the data that comes out, with patients declared ctDNA negative who were actually ctDNA positive but were not detected. We developed NeXT Personal to improve sensitivity by a factor of 10-100x.
Are some assays more appropriate for particular use cases or cancer indications?
Yes, the tumor-informed approach with higher sensitivity enables a move to an earlier stage of cancer, in a neo-adjuvant or adjuvant setting, and it also is more appropriate for cancers that have either a low tumor mutational burden (TMB) or a low shedding rate. One prime example of that, which is an area of high interest, is breast cancer, which is typically a low TMB cancer. We see traditional approaches where you might not have enough high-quality mutations to track accurately with these current methods, and that greatly limits your sensitivity.
Can you elaborate on the use of whole genome sequencing (WGS) for baseline testing compared to using panels or WES? Is this what is really driving increased sensitivity?
Our approach to drive increased sensitivity comes from sequencing the whole genome at the front end, using a tissue sample. If you look at the current tumor-informed approaches that use the whole exome, typically these approaches will track from 16 up to ~50 mutations depending on what method you look at.
The issue here is that, as you move to earlier stages of cancer, and as you move to low TMB cancers, sometimes it's not easy to find between 16 and 50 high-quality mutations to track, and that limits sensitivity. By moving to the whole genome, we can track mutations both in coding and non-coding regions of the genome. We do normal sequencing, as well, so we do not have issues with ChIPs. By looking across the entire genome, we can select 1,800 mutations to track. Not only that, but we also identify plenty of mutations and we can rank them to pick the highest frequency mutations – those that will give us a strong signal, as well as the ones that are in a low noise region of the genome from a sequencing standpoint.
We have high sensitivity also because of our advanced noise suppression in plasma. If you can pick regions with lower noise, that gives you an additional boost in sensitivity. We are tracking 1,800 mutations, so we are essentially casting a wider net in terms of our ability to detect ctDNA fragments in blood. If you’re tracking a ctDNA fragment in blood and you are only looking at 16 mutations, that's how many shots you have of finding that fragment. Instead, if you're looking at 1,800 mutations, you have at least 100x more chances to detect a fragment that carries one of these mutations in blood and that is where the increased sensitivity comes from.
What do you think will be the future interplay of therapy monitoring and therapy selection in the case of relapse?
One additional feature in our NeXT Personal assay is that we are not only looking at molecular residual disease (MRD), we are also tracking mutations at every time point along the journey to determine which mutations are at play when there is recurrence. Today, most solutions essentially provide a Yes / No answer (e.g., yes, you have cancer; or no, you do not have cancer). The feedback we’ve heard loud and clear from key opinion leaders (KOLs) was, “Okay, we can see cancer is recurring, but what should we do about it?”
That next level of information is missing with most assays, which is why we added tumor variant tracking to NeXT Personal. Every time we do a test, we report if cancer is present, Yes or No, and if it is present, we also report on what mutations are detected, by using a fixed panel. At recurrence, we can then identify what mutations are potentially at play in terms of actionable mutations or resistance mutations, which could then help inform what therapy should be utilized at that given stage of recurrence.
Finally, considering recent AGBT announcements, and excitement surrounding new platforms, would you like to comment on the novel platforms coming online (e.g., Singular Genomics G4, Element Bio Aviti)?
Our platforms tend to be on a larger scale, with our NeXT Personal assay based on front-end whole genome sequencing. Our platform has been built keeping in mind that we expect the throughput and cost of sequencing to decrease over time. As that happens, we think we can benefit compared to technology that is on a smaller scale (300-500 genes) because of the size of our platform. Engaging with new sequencing platforms such as Ultima Genomics, for instance, potentially gives us further capabilities to reduce the cost of our assays.
The premium in sensitivity within MRD remains. When we sit down with pharma and KOLs, this is the thing that attracts the most interest as the next step in MRD.