In 1989, Michael J. Fox reprised his role as Marty McFly for Back to the Future II (my favorite movie as a kid) and got everyone thinking about hoverboards and flying cars. Reagan bid farewell to the White House. Paula Abdul broke out with three hit singles. Seinfeld’s TV sitcom premiered. The Berlin Wall fell.
That same year Dr. George Savage and Andrew Thompson had a conversation that would eventually lead to the founding of Proteus Biomedical.
So the story goes, anyway.
Proteus, one of the original and highest-valued digital health companies, filed for Chapter 11 this week. This 4,400-word report digs into its mostly unreported origins, explores the company’s relationship with its biggest customer — Otsuka, uncovers the financial details of its recent deal with Tennessee’s Medicaid program, and considers what happens next for digital medicines.
Origins of Proteus Digital Health, née Proteus Biomedical
It was 1989 and the two Stanford business school students couldn’t get over an article they had read in Time magazine that explored applications for a technology called micro-electrical mechanical systems, or MEMS.
In his 2012 book, The Founder’s Dilemmas: Anticipating and Avoiding the Pitfalls That Can Sink a Startup, Noam Wasserman writes:
“In 1989, Andrew Thompson and George Savage, who had degrees in engineering and medicine, respectively, and had met at Stanford Business School, saw an article in Time magazine that described micro-electrical mechanical systems (MEMS). Andrew envisioned these as ‘little scissors that one day would float around in your bloodstream, gobbling up cholesterol’ and the duo was intrigued by the possibility of a startup. They concluded, however, that ‘while applications like these were potentially interesting, they couldn’t be built yet. We shelved the idea, but periodically asked ourselves, ‘is it time?’ George said, ‘Healthcare generally isn’t a good place to try out revolutionary technology that hasn’t been proven elsewhere.’ A decade later, however, Andrew and George decided that the technology to develop MEMS had matured to the point where the market for it might emerge quickly, sparking their founding of Proteus Biomedical.”
After graduating with Stanford MBAs and before starting Proteus in 2001, Savage and Thompson cofounded two medical device companies back-to-back. CardioRhythm and FemRx were each acquired by big medtech. Medtronic acquired CardioRhythm, which developed electrophysiological catheters to treat cardiac arrhythmias, in 1994 for about $26.5 million. And FemRx, which developed interventional gynecology devices, joined J&J following a $22 million acquisition in 1998.
Proteus has always focused on “silicon inside the body”
Proteus’ third co-founder and CTO was a seasoned MEMS expert prior to joining Proteus, Mark Zdeblick. Here’s how he described the company’s origins during an interview with the authors of the 2011 book, Edison in the Boardroom Revisited:
“In the early years before we were funded, we spent a long time exploring the different uses of silicon inside the body. We were specifically looking for implantable devices or electronics to add to existing therapies.”
The Proteus Biomedical of 2001 would be unrecognizable as the Proteus Digital Health we know today.
The company’s original “cornerstone” products were smart leads for cardiac resynchronization therapy (CRT) devices — pacemakers. Proteus sold these medical devices to CRT makers, including Medtronic, St Jude Medical, and Boston Scientific. Those early partnerships also explain why Proteus has so many big-name medtech companies as investors. While they may have bought into the intelligent medicine future that Proteus eventually focused on, it’s more likely they invested as a result of this earlier smart leads and related implantable device parts business.
The earliest iterations of the Proteus website’s homepage also make clear the company was focused on heart failure patients. The original Proteus tagline was “Revolutionizing Cardiology”.
2003: Thompson’s “stupidest idea” was Digital Medicines
One notable inconsistency in the Proteus origin story is its founding year. Some co-founders list the date as April 2001. Proteus’ recent Chapter 11 bankruptcy filing states 2002, making it the likely legal answer. Savage lists 2004 as Proteus’ start date. That may be because 2003 was the year the company’s founders came up with the idea for digital medicines.
Thompson told Forbes last year that he was attending the American Heart Association’s annual conference in 2003 when the idea for what would become Proteus’ Discover system came to him:
“[Thompson] noticed that none of the companies presenting at the conference were doing anything with digital technologies. ‘Silicon and software was innovating every other industry at the time except for pharma,’ Thompson says. So he brought his idea for a medicine that communicates with a computer to Savage. ‘George looks at me and says, ‘That is the stupidest idea you’ve had in your entire life.’ And then we began to argue,’ Thompson says.”
But Proteus’ revenue stream from making parts for cardiac device makers continued for many years after the concept of digital medicines entered the company’s business plan. Here’s Savage in 2009 explaining Proteus’ approach to the heart failure market:
“Our approach comprises two therapeutic technologies: implantable and ingestible. The implantable technology is being incorporated into CRT device leads as well as implanted devices for chronic pain and other disorders. Our ingestible technology is being incorporated into oral pharmaceuticals.”
Switching from implantables to ingestibles
While it’s hard to pinpoint exactly when Proteus’ business switched from a focus on implantable devices to a focus on ingestible ones, it’s clear that the original implantable device business was still driving Proteus in 2009.
The company remained focused on cardiology, and it saw ingestible sensors as another way to help heart failure patients take their meds. But it had more sweeping ambitions.
Here’s Savage in that same interview:
“The initial application for our ingestible therapy system is intended for the treatment of patients with heart failure. As suggested earlier symptomatic heart failure patients take multiple drugs daily to reduce the heart’s workload and improve quality of life. We are also investigating this ingestible system with tuberculosis therapy, as a less-invasive replacement for the standard of care, directly observed therapy, as well as psychiatric disorders, as a more effective way to connect customers with their caregivers. Ultimately, we envisage this offering significant benefit to anyone who has a chronic disorder and wants to more effectively and less invasively manage their wellness.”
One likely candidate year for Proteus’ transition away from implantables is 2012.
That’s when Proteus Biomedical rebranded as Proteus Digital Health to “better reflect” what the company does, as a Proteus exec explained to me at the time. In that same press release, Proteus announced it had inked a deal with Japan-based pharmaceutical company Otsuka. (I’ll dig into that more below.)
But first, it’s worth pausing to consider Proteus’ first decade of existence.
From 2001 until maybe as late as 2012, the company had a lesser-known business of supplying smart leads and other parts to implantable device makers. While this business, apparently, was not successful enough to propel Proteus fully in that direction, it seemed to be revenue-generating. After all, it helped convince big names like Medtronic and St. Jude to invest in Proteus during its early years.
In its Chapter 11 bankruptcy filing this week, Proteus says it was founded in 2002 “for the purpose of researching and developing Digital Medicines and, specifically, to improve outcomes for patients who regularly take prescription medication.” That statement isn’t true, as the above exploration of Proteus’ first decade of existence makes clear.
The filing also states:
“The Debtor’s business remains almost entirely in the ‘pre-revenue’ stage of development. Since inception, the Debtor has relied primarily on equity capital and advances under its agreements with Otsuka to finance its operations.”
Throughout its 19-year history, not all Proteus products were pre-revenue, but that does seem to be an apt description for Proteus today.
And as is likely obvious by now, the bankruptcy filing makes no mention of the company’s earlier implantable device parts business, which is assuredly now shuttered.
Otsuka: Proteus Digital Health’s one big customer?
When Proteus announced its deal with Otsuka in 2012, it wasn’t immediately apparent that this deal would become the startup’s most important one for the next eight years. According to Proteus’ Chapter 11 filing, a disagreement with Otsuka over the partnership would also trigger a series of events that led to the bankruptcy.
But first, a word about its first big pharma partnership: Novartis
Remember: Proteus already had a deal with a (much bigger) pharma company, Novartis, which invested in Proteus and secured a license to use Proteus’ technology in a 2010 deal worth $24 million. While the Novartis deal included an exclusive license for using Proteus technology in the area of organ transplantation, the pharma company also had certain rights to use it in cardiovascular and oncology therapeutic areas.
By the end of 2010, Novartis even said it hoped to put a Proteus-powered digital medicine through the regulatory paces in Europe within the next 18 months. That never materialized, however.
Even as Proteus signed its deal with Otsuka in 2012, I thought at the time that the partnership with Novartis was still the one to watch.
Novartis sponsored a small study of 20 organ transplant patients who used the Proteus system for 12-weeks. Results from the study landed in 2013. Remarkably, 40 percent of the group dropped out because of skin irritation from the wearable patch, digestive issues potentially from the ingestible sensor, or usability issues with the whole system. For those that stuck with the study, the system was able to accurately track the ingestion of the medication more than 99 percent of the time.
Still, Novartis never publicly shared any further progress on its work with Proteus. The study concluded that the wearable patch may be a problem for some people with sensitive skin, but the pharma company never explained why it let its Proteus partnership fizzle out.
I heard these wearability and usability issues along with an uncertain reimbursement environment — this was 2010, after all — led Novartis to get cold feet.
Novartis remained a fixture in Proteus’ long list of investors along with other past partners and customers like Medtronic and St. Jude.
Five different agencies within the FDA weighed in on Abilify MyCite approval
Over its 19 year existence, Proteus Digital Health crowning achievement was the FDA approval of Abilify MyCite. It took five years from Proteus signing the deal with Otsuka until the FDA announced the approval of the two companies’ digital medicine offering, which consisted of a wearable patch, companion app, and the pharmaceutical manufactured with an embedded sensor.
In April 2016, however, the FDA sent the companies a Complete Response letter, which asked them to make some changes to the app and to conduct some human factors studies. Finally, the FDA approved Abilify MyCite in November 2017.
Despite this five-year waiting game, Otsuka opted for slow, deliberate commercialization of Abilify MyCite
A year ago, Bill Zeruld, Senior Director, Alliance Management, for Otsuka Pharmaceutical Companies (US) explained in a podcast interview how Abilify MyCite’s launch had gone so far:
“Rather than a traditional big pharma splashy launch, we have been taking an extraordinarily deliberate approach. We received this approval in the fall of 2017. We have been reaching out and engaging with just a small number of plan partners, because we are basically taking the tack of this is different. This is not just another pill to prescribe. This is different.”
Abilify MyCite is available in some places today and it is listed in some popular compendia with an average selling price of $1,980 for a 30-pack, which includes 30 tablets with sensors and a patch.
Zeruld pointed to the complexity of the Abilify MyCite system as well as reimbursement issues as reasons to proceed slowly.
“There is the system that the provider needs to have in place because the physician needs to receive that data. How do they want to receive that data? How frequently? How meaningful is it and useful? As well as the reimbursement side of it. So, what are the studies — and we are doing real-world studies — that will be able to, we believe, demonstrate the impact that payers ultimately want to reimburse for. That’s a long way of getting to the point that we’re not just saying, ‘hey, we’ve got it, everybody, come to the party!’ We are working with partners to build this out in a very deliberate way.”
It’s just speculation but I think this fundamental difference in expectations around moving quickly vs. moving deliberately may have been at the heart of the issues Proteus faced in the end. Maybe pursuing a wide-scale rollout of Abilify MyCite instead of a smaller more deliberate one would have worked better than Otsuka anticipated and sent more revenue Proteus’ way. Or, maybe, Otsuka was right and most payers wouldn’t go for it without more data.
In any case, during many of the public talks that Otsuka executives gave about their work with Proteus, this tension came up. Proteus wanted to move faster and iterate more, while Otsuka pushed for slow and steady.
Here’s what Dr. William Carson, the President & CEO, Otsuka Pharmaceutical Development & Commercialization said in early 2019:
“We worked with Proteus to bring MyCite forward, and, early on, that was an interesting place for a pharmaceutical company to land, because the cadence in Silicon Valley vs. the pharmaceutical cadence is night and day. When we look at how to bring things forward, we say, well, that trial will be done in three years and we will be able to file [with the FDA] in five years. And on the Proteus side, it was: ‘how fast you can iterate?’ So, we had to, as an organization, work hard to make sure we were on the same page.”
During that same fireside chat, Kabir Nath, President & CEO, Otsuka North America Pharmaceuticals explained why it would take a long time for Abilify MyCite to get to scale:
“We also know that it’s going to take time to get to scale. It’s challenging, you know, with MyCite, we’re trying to change the behaviors of patients, physicians, and payors and many of us in the room know that doing any one of those things is pretty hard. Doing all three at the same time is extraordinarily hard. But we truly believe in the potential of this.”
Nath also built on that point by noting how hard it is to change behavior internally at pharma companies too:
“That’s a huge challenge because you’re right, it applies not just to timelines, but it applies to how we think about investment decisions. I often joke we’re very willing to spend $500 million on the toss of a coin, because any Phase III program in psychiatry is effectively a toss of a coin. Trying to get commitment for $10 million to move a digital therapeutics forward is much harder than doing the $500 million decision. And that’s not surprising because everyone’s comfortable with the zones that we’re in.”
In a different interview from May 2019, Nath made clear that Abilify MyCite (and other digital medicines that Otsuka would create with Proteus technology) still had a long road ahead of them:
“We want to actually start to build the evidence that we are measuring ingestion, [that] it does actually improve adherence, [that] it does improve outcomes and get to the point where we’re working with payers on outcomes-based [reimbursement]. We have to generate that evidence. That’s still the vision. It’s still going to take some time. We’re still some distance from the tipping point.”
Did a slower-than-expected Abilify MyCite rollout lead to Proteus Digital Health’s bankruptcy?
Before we get to the most interesting paragraph in the Proteus Chapter 11 filing, it’s important to note that Proteus and Otsuka expanded their agreement in 2018. The new $88 million deal included a broader licensing arrangement that would lead to the creation of a portfolio of antipsychotics coupled with Proteus technology. The companies called this the Expanded Collaboration Agreement.
Here’s that paragraph from the filing:
“Leading up to the fourth quarter of 2019, [Proteus] experienced a severe liquidity crisis while attempting to resolve certain issues with Otsuka in connection with the Expanded Collaboration Agreement. The liquidity crisis prompted [Proteus] to issue WARN notices and then furlough nearly all of its 300 employees during the last week of October and for the first two weeks of November 2019.”
I can read that paragraph more than one way, but it sounds like the “issues with Otsuka” were the cause of this “severe liquidity crisis”. If they weren’t, why mention those issues here?
By the end of 2019, Otsuka and Proteus renegotiated their agreement yet again. Otsuka agreed to a series of payments to Proteus that totaled $90 million, which were timed to allow Proteus to pay down its debt to OrbiMed with enough leftover to keep its operations running.
Proteus realized that it would still need additional funding to stay afloat in the first half of 2020. Despite attempts to raise money, the onset of the COVID–19 pandemic in early 2020 “created significant uncertainty in the capital markets and frustrated” Proteus’ efforts.
Proteus’ final move was to hire a banker and attempt to sell the company before it ran out of money. While it managed to distribute confidential memos to 29 parties under NDA, Proteus’ banker received no bids by the June 10 deadline. That’s when it filed for Chapter 11.
Otsuka breaks up with Proteus in early 2020
In the middle of all that, Otsuka and Proteus made another agreement to help Otsuka transition out of the partnership. E&O covered this part of the story in Issue 035 in January 2020:
“Proteus has ended its deal with Otsuka, which resulted in the biopharma paying out Proteus to license the technology they built together for mental health therapeutic areas. Proteus will use that payment to stay afloat for now.”
The breakup included at least one patent, which Proteus assigned to Otsuka following this deal. Otsuka also brought on some of the team at Proteus that worked on the project with them, including one of Proteus’ earliest employees, Todd Thompson, who joined Otsuka as its VP of Devices and Technology in March.
Details on Proteus’ final customer win: Tennessee’s Medicaid program, TennCare
This move is a bit of a headscratcher for me. I understand that Proteus needed cash, but this deal has to be too little, too late, right?
In May 2020, just weeks before filing for Chapter 11 bankruptcy, Proteus Digital Health announced Tennessee’s Medicaid program, TennCare, as its first new customer win in more than a year.
The outcomes-based contract will provide:
“TennCare-covered patients undergoing Hepatitis C treatment to achieve a cure through improved medication adherence and stronger connections to their care teams. Under the terms of this contract, Proteus will only receive payment from TennCare if a patient on digital medicine therapy successfully completes treatment and is cured.”
I managed to find the contract: Here’s how the payment will work (or would have worked?):
And, surprisingly, the total possible payment for this “pilot program”, which maxes out at 250 patients, probably wouldn’t even cover the average salary for one Proteus employee in 2020. The total funding for the pilot is as follows:
Was Proteus Digital Health too focused on Otsuka? Or too unfocused overall?
The prevailing armchair analysis of “what went wrong at Proteus” these past few months is that the company should not have focused on schizophrenia as its lead therapeutic area. Even late-night TV hosts like Stephen Colbert joked about Proteus’ plans to sell spyware-embedded meds to people with schizophrenia.
Hard to disagree here. It’s also hard to calculate what becoming a running joke in your own industry does to your brand equity and partner prospects when you are still “mostly pre-revenue”. Did it matter? All press is good press, maybe?
Over the course of its 19-year run, however, Proteus spent time developing a wide range of products — not just Abilify MyCite. How many are too many? At one point do dead-end products and pilots start to drag on the morale of a startup? How different is a digital medicine for one therapeutic area versus another?
- In 2004, the earliest product that Proteus vaguely described in a press release was the first version of its medical devices for heart failure patients: “Proteus’ early-stage platform includes a catheter, lead system, and an implant.”
- One of Proteus’ earliest products was ChipSkin, “an extremely thin and durable wrapper to protect the chip and preserve the performance of the ICs in the otherwise inhospitable environment of the human body.”
- By 2008 Proteus had product descriptions for its first two implantable device products: The first one addressed the clinical issue of extra-cardiac stimulation, which affected 10-20 percent of CRT implants at the time. The second used what Proteus called “electric tomography”, a proprietary real-time modality for quantifying cardiac performance.
- In 2008 Proteus also referred to its ingestible event market sensor as its Raisin technology, “which transforms existing drugs into intelligent medicines with the simple addition of a tiny ingestible microchip to a capsule or tablet during final product manufacturing, requiring no alteration to drug formulation.”
- In 2010 Proteus announced an exclusive deal with its new investor, Novartis, for digital medicines for organ transplant patients.
- In 2011 Proteus revealed its metabolic-focused digital medicines program, which it called Equa. Equa was a digital health program built around taking metformin. It hoped to launch it in 2012.
- In 2011 Proteus also talked up a suite of digital health programs, called the HealthTiles Operating System. It offered a customizable mix of programs for weight loss, sleep, medication reminders, and more.
- In 2011 Proteus began an NGO-funded, Proteus technology-powered case management program in China for patients with tuberculosis.
- In 2011 Proteus had a licensing deal in place with Avery Dennison to supply it with its wearable, peel-and-stick patch technology for various consumer and healthcare remote patient monitoring applications.
- In 2012 Proteus researchers presented the results of a small study that investigated a gamified, sensor-enabled mobile-based health program Proteus created called the Power Challenge Personal (PCP).
- In 2012 Proteus inked its first commercial deployment of Helius, the wearable patch, ingestible sensor (swallowed alongside a pill in this configuration), and companion app. The deal was with a chain of pharmacies in the UK called Lloyd’s. It was a direct-to-consumer play.
- In 2012 a Proteus presentation describes Helius-powered programs for uncontrolled hypertension, “transition to home”, diabetes, and dementia.
- While Proteus had mentioned “MedMatch” in passing for a few years, in 2013 Fast Company made this dubious claim about this health game that Proteus created: “That’s why many drug companies signed on to participate in Proteus’s new program, MedMatch: When patients take their pills regularly, the companies donate drugs to people who can’t afford them. ‘Patients like it because instead of the drug company controlling them and saying, You have to take your pills, patients get to say, You have to give away your pills,’ Thompson says. And everyone gets healthier.”
- By 2016, one medical journal reported that 400 different drugs had been co-ingested with Proteus’ standalone ingestible event market sensor.
- In 2018 Proteus announced that it had 31 digital medicines in its pipeline. Some 15 digital medicines for cardiovascular and metabolic conditions were available, along with digital meds for infectious diseases like hepatitis C, HIV, and TB. It also said it was developing digital meds based on oral meds for cancer patients and drugs intended to help with side effects of cancer treatment, including opioid analgesics.
- In its 2020 bankruptcy filing, Proteus claims to have more than 20 digital medicines in its product portfolio.
I’m sure there are other products and concepts that Proteus devoted resources to that never made it into a company presentation or Fast Company article.
So, while many will say that Proteus focused too exclusively on schizophrenia, it’s not as if the company wasn’t actively trying to launch other products. In fact, it may be that Proteus tried to bring too many products to the market at the same time.
What happens next for digital medicines?
The fate of Proteus
While unlikely, there is an outside chance that Proteus restructures and soldiers on in some new, pared-down form. Of course, that would be with a new set of owners and a new executive team. Maybe this leaner Proteus would pursue smaller deals like the one its predecessor inked with TennCare. Maybe it’d become more of a partner to other digital health companies and compete with EtectRx head-on (more on that below).
I’d guess that two other scenarios are more likely: Either Otsuka acquires some of Proteus’ assets or — considering its hundreds of patents — a patent troll does. At this stage, however, I imagine if Otsuka was willing to make that move they would have already. Curious how this one shakes out.
Otsuka seems to be willing to take digital medicines in-house, for now
As noted above, Otsuka hired some of Proteus’ talent earlier this year and it acquired some of the company’s IP in order to keep Abilify MyCite, at least, in the market. If Otsuka does not acquire the rest of Proteus’ assets, how will it continue to manufacture its digital medicines? Were those transitionary moves part of a longterm strategy or just a way to gently let its digital medicines program sunset? They may have been defensive moves to protect their investment in the Proteus project so far while they await the bankrupt company’s outcome.
A major opening for EtectRx
One of Proteus Digital Health’s only competitors, EtectRx, managed to secure FDA clearance for its ingestible event marker (IEM) a few days after news of Proteus’ financial troubles surfaced in the press. I dug into the different strategies of EtectRx and Proteus in Issue 029 last November:
“Both Proteus and EtectRx have developed smart pills that send signals from inside the patient’s body when they come into contact with gastric fluids in the stomach. In their current configurations, however, where that signal goes differs.
“Proteus sends its ingestible sensor signal through the patient’s body (which is the conductor) to a patch worn on the patient’s abdomen. This patch collects other biometrics as needed and relays the ingestion signal along with other biometric data to the patient’s smartphone.
“Instead of a peel-and-stick patch, EtectRx currently has the patient wear a receiver around their neck, but the company hopes to send that signal to a variety of devices in the future. That’s why it wants to standardize smart pills’ communication protocol.”
While it still has a long road ahead of it, EtectRx might manage to avoid at least some of the wearability and usability issues that probably tanked Proteus’ deal with Novartis. By avoiding the peel-and-stick wearable and finding a way to send a signal from its IEM directly to the patient’s phone or smartwatch, EtectRx might avoid Proteus’ fate.
The other big strategic difference is that EtectRx is willing and focused on partnering with other digital health companies. One thing that is striking about Proteus — arguably one of the first digital health companies — is that it never partnered with other digital health companies. If EtectRx succeeds in its current strategy, it will partner with pharmaceutical companies, digital therapeutics companies, disease management companies, and more.
Reimbursement for digital medicines is the last nut to crack
In the 2019 podcast interview linked above, Otsuka’s Bill Zeruld stressed that reimbursement remains the key hurdle for digital medicines:
“It’s still early, but our optimism remains extraordinarily high… I would call it cautious optimism. Cautious because I would say we haven’t seen a laundry list of successful business models… Reimbursement seems to be the nut that has not been fully cracked just yet. So, what is the business model? There may be a few out there who can tout some success, but I’d say they are the exception, not the rule. Until we get there, I am going to remain cautiously optimistic about it… so, we are still early. Even though this has been going on for a few years, we are still just on the cusp.”
