Perimeter Medical Imaging AI: Covering the Margin Calls for Cancer

07/29/2020

For the last eight months, the Covid-19 virus has been the centre of attention in the media, the markets, and the medical field. Collectively, countries, government agencies, academic institutions and pharmaceutical companies across the world have poured hundreds of millions into efforts combating the viral outbreak that has left so many countries and populations struggling during their wait for a vaccine. That’s the key word there though: a vaccine. Though Covid-19 rightfully dominates the global attention at the moment, it’s cancer that remains the most daunting C-word in the medical arena because of our current inability to realistically imagine a future where it could be preventable by a vaccine.

We often say “we’re not curing cancer here,” – but what if we were giving it a shot? Or at the very least, making its treatment conceivably more effective by taking massive strides, with technology from an unprecedented angle, that is seen as important enough by those in the know that over $10 million has been granted to help get it done, with $10 million more raised in the last few weeks?

To be clear, we’re not claiming that this is some kind of spectacular company that’s come out of the blue to just drop in your laps – this company with the industry-changing tech went through the right hoops, is backed by the right people with the right credibility, but has simply strategically sailed under the radar until now. They’re ready now though.

 

Perimeter Medical Imaging (PINK.V) recently landed on the public markets and, after some early position escaping it got ripped down from a $1.70 to $1.50 in the first few days (for reference, PINK.V was originally $1.50 at open), is starting to find its feet at that level.

The story isn’t as easily told as your average weed grow or mushroom spec play, so please bear with us. Medtech isn’t traditionally sexy in Canada (or really anywhere) because the long wait for FDA approval and the requisite waits for confirmed breakthroughs results in slow months for news – all things much harder to market and incur regular interest than a catchy “let’s get stoned together with [new company X]!”

We feel however, that medtech can be a more rewarding industry to invest in for some – those who take the time to learn the details of the science and the standard timelines. If you do the research, you could pick some wow-time winners with the added benefit of being invested in things that can leave the world a better place.

The Covid-19 investment wave has changed some of the medtech focus, with first-time investors in the space getting drawn in because of its virality. However, Covid-19 has shown tentatively promising results for vaccines already.

Cancer doesn’t go away with a vaccine – and unlike Covid-19, a critical mass of mask-wearing isn’t going to significantly and reliably reduce your susceptibility for it.

Covid-19 is hitting hard, but cancer has been hitting hard for centuries because there’s simply no hiding.

 

A quick run-through

It’s a fact of daily life that we are all regularly exposed to cancer-causing agents such as sunlight, radiation, environmental chemicals, and substances in the food we eat. These agents cause mutations in our DNA and, as we age, our immune system becomes less efficient in dealing with cells containing that mutated DNA. Therefore, advancing age is the most important risk factor for cancer overall, which also means if you’re not at risk so much today, tomorrow is going to be different. As life expectancy increases globally, so do cancer numbers, devastating families every day.

 

Source: https://www.ncbi.nlm.nih.gov/pmc/articles/PMC6530937/[/caption]

 

“Cancer is the second most common cause of death after ischemic heart disease, according to World Health Organization (WHO) estimates. Among women aged 30–59, cancer is the leading cause of death… Projections made by WHO in 2013 predict that by the year 2030 cancer will surpass ischemic heart disease and become the most common cause of death, accounting for 18.5% of all deaths.”

The types of cancer responsible for deaths in men are predominately lung (23%), liver (11%), stomach (9%), colon (8%), and prostate (7%). In women, breast cancer predominates (15%), followed by lung (13%), colon (9%), cervical (7%), and stomach (also 7%). These data are summarized below.

Recently listed on the TSX venture exchange as PINK.V – an homage to the pink ribbons used during Breast Cancer Awareness Month by the Canadian Cancer Society and the American Cancer Society – Perimeter Medical Imaging AI, is a Toronto and Dallas based company dedicated to providing solutions that drive better patient care and lower healthcare costs by providing critical information during clinical procedures.

That’s the corporate talk, at least.

 

Translated: “Perimeter develops, patents, and commercializes advanced in-procedural medical imaging tools. The OTISTM platform provides a crucial additional dataset to the clinician to facilitate more informed decision making that could help reduce over-sampling, lower the incidence of repeat procedures, and potentially even lower mortality rates.”

In layman’s terms, Perimeter’s OTIS platform is designed to achieve the goal of providing surgeons, radiologists, and pathologists with real-time, ultra-high resolution images which give them the ability to review tissue microstructure during surgery.

 

Still too complex? I got you.

When it comes to removing a tumor, your surgeon faces a delicate balance. Clearly, there’s nothing more important than removing all traces of the offending tumor. At the same time, surgeons don’t want to scythe out large areas of tissue like something out of a Texas Chainsaw Massacre deleted scene. It’s simply not done (or nice, to put it mildly).

Once a tumor is removed, the area around the removed tumor is known as ‘margin.’ All of the margin is then tested to check for traces of cancerous tissue.

If the tissue is declared Clear (also called Negative or Clean), it means no cancer cells are seen at the outer edge of the tissue that was removed. If the tissue is declared Positive, cancer cells are present right out to the edge of the removed tissue, indicating the tumour’s reach was larger than expected and more surgery is usually needed – with all of its expected risks and complications.

 

According to breastcancer.org:

“In a perfect world, your surgeon would learn the status of the margins before the lumpectomy (removal of cancerous breast tissue) is completed, so as much tissue as needed can be removed until the margins are clean. Unfortunately, analyzing the removed tissue takes about a week. Sometimes after the pathology report is done, the margins are found to contain cancer cells and more surgery is needed”.

Now, for the first time hopefully, put yourself in the shoes of a cancer surgery patient. You’ve already been under the knife once and you’re living on the edge of hope and more-chemo, wondering if they got everything and whether you can call yourself a survivor now. Instead of good news, you get a call from the doctor saying, “oops, little bit left, let’s get back in there.”

Enter Perimeter, which has the OTISTM system in place to hopefully take us one step closer to breastcancer.org’s perfect world. With the Perimeter system, your surgeon can visualize the margins during the surgery, which means, your surgeon could check for unexpected cancerous material as (s)he goes, and clear everything that needs cleaning out – the first time around.

As Perimeter and its OTIS system navigate through the intricacies of the FDA process, let’s give you an idea of how that generally works, as an FYI.

In 2019 OTIS platform received a 510(k) clearance.

 

What is a 510(k) clearance? 

A 510(k) is for devices that are substantially equivalent to other 510(k) devices. FDA clears these products for marketing but does not approve them. Approval is limited to a more detailed review of product safety and effectiveness for Premarket Approval (PMA) devices. In 1976 when the device law was enacted, FDA took everything that was in the market and put it in one of three categories: Class I, II or III.

Class I and II products are those that were eligible for 510(k), while Class III were allowed to remain in the market until FDA called for PMAs. What this has done over the course of the years is to attenuate somewhat the comparisons, since it is a little difficult to compare a 2020 device to something that has been in the market since before 1976.

So, to avoid this situation where we are comparing a TESLA to a 1950’s Chevrolet Blaire, Congress changes the law to say that you could compare the device to something that was lawfully on the market under a 510k in order to show substantial equivalence.

There is one additional category that the Congress created in the early 2000’s: the De Novo 510(k) is limited to products for which there are no predicates but which are very low risk. Without the De Novo process being available, companies would have to do a PMA for devices of that type, which didn’t make sense if the product was truly low risk. The De Novo category allows products to be marketed under a 510(k) as long as the company can demonstrate to FDA that the product is low-risk and that they can develop special controls to meet the requirements of a 510(k).

The perfect example of this would be Perimeter’s ImgAssist AI technology, where the device is given data of lumpectomy surgery margins to decipher the data and uses its AI capabilities to alert the surgeon of areas within the margin that are suspected of containing cancer cells.

Now let’s put that into perspective in terms of what Perimeter has and where it is in comparison to what’s already available in the market.

OTIS’s commercial platform with 10x improved imaging speed, proprietary tissue handling system and image atlas achieved US FDA clearance in March 2019. OTISTM product is cleared under general label, which allows it to take sub surface images (image tissue out of the body).

AI-based breast lesion detection algorithm hopes to receive clearance via the US FDA De Novo pathway. Even better, the algorithm will be fed data collected from various devices installed across the partnering clinics.

 

This leads us into Perimeter’s last press release:

 

Dallas, TX – July 15th – Perimeter Medical Imaging AI (TSX.V: PINK) today announced the initiation of its ATLAS AI project that will collect images of breast tumors to both train and test their ImgAssist AI technology which is currently under development. This technology is designed to utilize a machine learning model to help surgeons identify, in real-time, if cancer is still present when performing breast-conserving surgery (“lumpectomy”). Perimeter will install OTIS™ high-resolution imaging devices at their partner sites, and patient recruitment will begin immediately with the expectation of enrolling up to 400 patients.  This study was made possible, in part, by a $7.4 million grant awarded by the Cancer Prevention and Research Institute of Texas (“CPRIT”), a leading state body funding cancer research.

 

The first phase of the project will focus on data collected from leading pathology labs in Texas: MD Anderson, Baylor College of Medicine and UT Health San Antonio. Specially installed OTIS™ devices will collect image data of breast tissue samples from approximately 400 patients to refine ImgAssist AI- the proprietary software model using machine learning, which is adept at interpreting imaging data when labeled effectively.”

 

What’s next?

As with any AI platform, the more data that’s available from which to learn, the more likely the algorithms are to determine the correct outcome. Based on that logic, Perimeter’s goal should be to install OTISTM devices in more partnering sites, collecting more data.

As OTISTM comes into more use, and the results flow in regarding margin, and patient cases are updated through ImgAssist AI, Perimeter’s proprietary technology will have the benefit of having compounded AI learning on data. While Perimeter’s business model is solid, and has steadily achieved results, it does have competitors, and Perimeter’s scope of long-term success will be dependent on how their technology performs in comparison.

That said, here’s a standard but deeply necessary disclaimer to acknowledge that along with its competitors, Perimeter as a medtech company runs the risk of getting hung up on testing, partnerships that go slower than hoped, trials that never materialize, and the like. However, there’s little risk Perimeter will run out of cash in the short term, which takes care of that concern.

Ultimately, Perimeter’s success is dependent on the reliability of the data it collects, which primarily depends on achieving a critical mass of partnering clinics, their only true main risk. After all, a proprietary AI is only as good as its learning material.

 

SHOULD I INVEST IN THIS COMPANY? 

So if we’re going to invest in something that we hope obliterates the need for operation re-dos, we should figure out how often those even happen.

And unfortunately, oh, do they happen.

About one-quarter of Canadian women who have a “lumpectomy” to remove a cancer tumour will have a second breast operation. The 25 percent figure mirrors averages in the United States and Europe.

Frankly, the mental anguish alone of going through another surgery to remove even more cancerous tissue is one that no patient should ever experience. Additional surgery means medically, additional chance of complications as well as potential infections.

Financially, it might mean even more. The second surgery is costly. In the US, lumpectomy re-operations cost the medical system (and by this, we mean the patients) $560M USD annually.

 

Locally, numbers in BC are even worse.

The rate of re-operations in B.C. is 37 percent according to statistics collected by the Canadian Institute for Health Information.

Think about that now. 37 women out of every 100 that go under the knife, come back for a second incision weeks later, bringing their families with them in this nauseatingly hopeful journey.

Perimeter Medical Imaging AI’s system could potentially make all of that unnecessary.

From a market perspective, there’s nothing objectionable about how the company came to market. With a very tight cap structure with only 38,680,773 shares issued, you’ve got to imagine there’s potential here for upward stock traction, if even solely on the back of the news that confirms the system is being used for its intended use.

What we really like is though, is that breast cancer treatment (as massive a market as that is) isn’t the ceiling on use of this technology. As long as they can haul in the data for their AI system to churn through, there’s potential here for use with any surgery that has similar earmarks for cancerous margin.

It’s no secret that we like making money. If we can make money with something that literally would make the world a better place, we’re all in on that.

 

 

Related Posts

Latest Post

Leave a Reply

Your email address will not be published. Required fields are marked *