Friday, July 26, 2019

When tech entrepreneurs attack science

One of the most remarkable trends in science-based entrepreneurship is the recent explosion of fake meat companies.

I saw my first fake meat company in 2015 at the graduation event for the first class of startups at the IndieBio accelerator in San Francisco, doing egg whites. Now leading companies like Impossible Foods and Beyond Meat have landed their fake hamburger in fast food chains.

At breakfast yesterday with my boss’s boss, he remarked that some of the burgers are actually quite good, and we agreed that (someday) it has the potential to be a trillion dollar business. This seems to be one of the rare examples where the outrageous predictions by tech entrepreneurs of creating a huge new market might actually be true.

Adoption Paths

The initial pioneers may grab decent exit values, and the long-term future of replacing meat seems compelling. California alone spends 6 trillion liters of water a year on alfalfa alone (feed for cattle and horses), not counting other states, and the water for pig slop and chicken feed. Meanwhile, cow farts play a non-neglgible role in increasing greenhouse gasses. And there is also a sizable niche of the populace that either refused to eat meat, or even wants to deny others the right to do so.

It’s not clear when it will become a trillion dollar market: as I showed in my 2014 paper in the Journal of Technology Transfer, California firms created the solar industry but flamed out because they got into the market 20-30 years too early. Competing with commodity electrons is a tough adoption curve: very few people voluntarily choose to pay 50% or 100% more than market prices for a commodity, although Germany and California show that politicians can force their voters to do so and (mostly) get away with it.

What I didn’t realize until I thought it through is that meat has an easier adoption curve, with a wide range of niche markets that can be sustained at premium prices. You have affluent people who don’t eat meat — or, even better, recently gave up meat — as well as environmentally conscious customers who would like to avoid meat. You have people who are willing to give it a try, out of curiosity. And — as the burger joints have demonstrated — the B2B customer (distribution) is willing to try a niche product to raise average selling prices.

Thus, as the product gets better and the prices get lower, these firms can establish and grow their beachhead in the food market, carving off ever-larger segments of the market. Funded by Sand Hill Road and led by ambitious entrepreneurs, some will hold off for Facebook-style IPOs, but many of the weaker players will be bought up by ADM, ConAgra, Hormel and the like — providing bottomless capital to spur innovation and adoption. (The entry barriers are low enough that Tyson Foods is launching its own product directly, rather than by acquisition).

We Need Science

However, to fully displace meat, there are major technical challenges to be overcome, both in quality and cost. I supervised a student project to research synthetic organs — a more demanding applications — but still getting the texture right will require both science (new insights) and engineering (new applications) to create a quality product at a competitive price.

Thus, I was struck by the decision of one fake meat company to attack GMOs to win market share. Yes, the CEO is a 24-year-old recent Berkeley grad who’s never worked in a company. Yes, her bachelor’s degrees are in toxicology and environmental studies rather than molecular biology or chemical engineering. But the company does have one PhD (food science) in its leadership, so they presumably are doing actual science.

It reminds me (and not in a good way) of the various surveys that showed the gap between what the public thinks and what scientists (writ large) think about GMOs, including a 2015 survey that said 37% of the public thought GMOs are safe vs. 88% of scientists.

More troubling is that the certainty of these opinions seems inversely proportional to actual knowledge. As the NY Times wrote on Monday:
In a paper published early this year in Nature Human Behavior, scientists asked 500 Americans what they thought about foods that contained genetically modified organisms.
The vast majority, more than 90 percent, opposed their use. This belief is in conflict with the consensus of scientists. Almost 90 percent of them believe G.M.O.s are safe — and can be of great benefit.
The second finding of the study was more eye-opening. Those who were most opposed to genetically modified foods believed they were the most knowledgeable about this issue, yet scored the lowest on actual tests of scientific knowledge.
In other words, those with the least understanding of science had the most science-opposed views, but thought they knew the most. Lest anyone think this is only an American phenomenon, the study was also conducted in France and Germany, with similar results.
So I get that trust in authority has been declining since the 1970s. I get that we have many people who don’t understand — or have the time to personally verify — scientific research. And, as Orwell predicted (and Goebbels proved), people are easily persuaded to believe lies if they are repeated often enough in the mass media.

Still, why would companies that depend on scientists to create their products help promote such lies? Isn’t the benefit of saving the planet enough, without having to rely on junk science for the purpose of virtue signaling? And if companies that depend on science attack science, what are the implications for K-12 and university science indication, science policy and the idea of using facts — rather than emotion - as the basis for making science policy?

Monday, February 13, 2017

Complexity, risk and uncertainty

We unexpectedly lost a close family member over the weekend, who was in the ER and then had such a massive internal hemorrhage that they couldn’t save him. The ER doctor said it wouldn’t have been as severe — and implied that he might not have died — if he hadn't been on warfarin.

All weekend I have been thinking that while the risk of bleeding is a well known side effect — one he and his doctor willingly accepted — if only he hadn’t taken this medication we would still have him with us.

But then this morning I realized I had it all wrong. As an economist, I should know better — this is just another version of Bastiat’s broken window fallacy. We can see the death by internal bleeding at age 86; we can’t see the heart attack or (worse yet) disabling stroke that might have been prevented at 80 or 84.

This highlights the hubris of mankind’s attempts to understand the most complex organisms in creation – ourselves. Yes researchers try and should continue to try to do better, to extend life and continue to reduce the mortality and morbidity of major illnesses. But nonprofits using marketing slogans like “abolish cancer” are dishonest at best and fraudulent at worst, following in the footsteps of the politician who 45 years ago promised to win a “war on cancer” to distract from his losing a war overseas.

The reality is that we are generations if not centuries away from really understanding how it all works and how to prevent or repair many common fatal or disabling diseases. The doctors on the front lines— like 18-year veteran I met in the ER Saturday night — are painfully aware as to the limitations of what we know. In the meantime, there are trade-offs and risks in life — in clinical trials, with new drugs, with old drugs — and all we can do is improve the odds rather than hope to know what is “best” in any given situation.

Wednesday, November 2, 2016

Pricing is only a symptom of drug company woes

Drug companies have been in the news a lot this year. Valeant and Mylan (with the EpiPen) made themselves whipping boys (or girls) through their self-inflicted PR debacles. Next Tuesday, California citizens will vote on an initiative by anti-pharma activists (aided by a failed presidential candidate) who seek price controls for 12% of the drug purchases in the state.

This morning’s WSJ reminds us that none of this controversy changes the stark reality of the drug business: since the Hatch-Waxman Act was passed in 1984, most* drugs have a narrow window to generate profits from their R&D before they lose their pricing power due to a flood of generic competition — and for the largest product categories, they will face competition prior to patent expiration.
Heard on the Street

Big Pharma Sales Need a Booster Shot​

By CHARLEY GRANT

Wall Street Journal, November 2, 2016, C14

The coming election has pharmaceutical investors feeling anxious. But this earnings season has highlighted a more solid reason for worry about the sector: Sales of key blockbuster drugs are slowing down.

This year has been one to forget for the industry. Political risk surrounding the high cost of health care has contributed to the trouble: The Nasdaq Biotechnology Index is down nearly 30% since a furor erupted over high drug prices in September of last year.

Price pressure has been among the reasons sales are weaker.
The article lists a number of examples of such competition:
  • More products (and thus greater price competition) for hepatitis C treatments
  • Disappointing revenue growth for Ibrance (breast cancer) and the Humira and Enbrell anti-inflammatories
  • Slowdown in diabetes sales
  • Price resistance for new cardiovascular drugs
It then concludes:
A basic truth of drug development is the nature and timing of economically significant breakthroughs can be hard to predict. Today’s slow patch doesn’t mean that the industry has become worse at developing blockbuster drugs.

But investors tend to prefer companies with easily predictable growth prospects. And it isn’t clear what new drugs can generate growth in the near future.

Even in areas where there have been discoveries, the outlook may not be great. A number of companies have developed impressive immunotherapy drugs to treat cancer, but that is a crowded field with lots of competition. And since these drugs are new, just how many cancer patients this class of drugs can help isn’t yet known.
In other words, pharma faces the same challenges they have always faced: attractive markets attract competition, buyers exploit competition to reduce pricing power, and then they have to develop new compounds to replace those going off patent. If a firm doesn’t address such problems year in and year out, its revenues, profits and market cap will collapse. We needn’t cry for Big Pharma, but neither should we underestimate the magnitude and complexity of their challenges.

In many ways, the free market works better for expensive drugs than for healthcare services. Payers and providers are organized enough to bargain for better prices from manufacturers, and the overhang of generics (the “Better than the Beatles” syndrome) forces manufacturers to produce markedly better outcomes to justify proprietary prices.

Meanwhile, in the U.S. our third party payer system (a side-effect of WW II wage controls), provider networks and high switching costs mean that patients rarely have pricing information, quality information or purchase alternatives to bargain based on price or quality. In single-payer countries, patients have less choice than public school students (who at least have charter, private or home school alternatives).

*PS: For technical and regulatory reasons, large molecules do not yet face generic competition because biosimilars are not generics. It seems likely that this will eventually change, but for now an expired patent on a biologic is worth a lot more than an expired patent on a small-molecule drug.

Tuesday, March 8, 2016

UC's biggest drug bonanza - ever?

On Friday, UCLA announced it had sold rights to the prostate drug Xtandi (enzalutamide). The sale could make the drug University of California’s most valuable biomedical patent family — ever.

UCLA, the inventors and its partners will receive $1.14 billion cash (plus future payments) for the drug from Royalty Pharma. This IP investment company owns stakes in various blockbuster drugs, including Humira, Remicade and Lyrica. It is Royalty’s biggest deal since the $3.3 billion it agreed to pay in November 2014 for royalties on Kalydeco.

In August 2005, San Francisco-based Medivation licensed Xtandi from UCLA and received US regulatory approval in September 2012. Medivation relies on Japanese pharmaceutical giant Astellas Pharma to distribute the drug worldwide. Its 2015 10-K states
Under our collaboration agreement with Astellas, we share equally with Astellas all profits (losses) related to U.S. net sales of XTANDI. We also receive royalties ranging from the low teens to the low twenties as a percentage of ex-U.S. XTANDI net sales. 
The drug has generated more than $3.4 billion in sales through December 2015. According to the 10-K statements, the global sales totaled $1.9 billion in 2015 and $1.06 billion in 2014. Its US sales were $392.4 million in 2013, and its US (i.e. only) sales in 2012 were $71.5 million.

Medivation’s 10-K states
We are required to pay UCLA (a) an annual maintenance fee, (b) $2.8 million in aggregate milestone payments upon achievement of certain development and regulatory milestone events with respect to XTANDI (all of which has been paid as of December 31, 2015), (c) ten percent of all Sublicensing Income, as defined in the agreement, which we earn under the Astellas Collaboration Agreement, and (d) a four percent royalty on global net sales of XTANDI, as defined. Under the terms of the Astellas Collaboration Agreement, we share this royalty obligation equally with Astellas with respect to sales in the United States, and Astellas is responsible for this entire royalty obligation with respect to sales outside of the United States.
UCLA and Medivation have had at least two lawsuits over the terms of this agreement. According to the 10-K, UCLA has accused it of not paying the 10% sublicensing royalty. Earlier, Medivation sued (unsuccessfully) to block UCLA’s licensing of a related compound to Aragon Pharmaceuticals.

UCLA received $33 million for multiple prostate cancer patents in FY 2013-2014 — the highest in the University of California that year. But 4% of $3.4 billion thus far should be worth about $136 million beyond the $1.14 billion for a total of $1.275m. This would not include “potential additional payments” from Royal Pharma, or revenues from the Aragon license.

As best I can tell, the most lucrative patent in University of California history. Many of us assumed that the previous winner was the family of three Cohen-Boyer patents, which created the biotechnology revolution through recombinant DNA, and allowed Herb Boyer to co-found Genentech. These patents expired in 1997; the best estimate I’ve seen (Feldman et al 2007) places the total licensing revenues from those patents at $255 million, split between Stanford (Cohen) and UCSF (Boyer). This total does not include the $300 million that Genentech paid (after years of litigation) to City of Hope — a LA-area cancer research hospital — for related discoveries.

Finally, what is often not remarked is that at most US universities, royalties are shared with the inventors. The inventors in this case Michael Jung of UCLA,  Charles Sawyers (then a Howard Hughes Medical Institute researcher at UCLA, now at Sloan Kettering). The USPTO lists 8 granted patents jointly authored by the two men, including two (8,445,507 and 8,802,689) explicitly about prostate cancer therapy.

According to the UCLA patent policies, UCLA keeps 50% of patent royalties, 35% goes to the inventor and 15% goes to the inventor’s lab (which for Jung is the chemistry department). The Royalty Pharma press release says:
By virtue of patent and licensing agreements administered by UCLA, the campus, the researchers and Howard Hughes Medical Institute shared a royalty interest in worldwide net sales of Xtandi. UCLA owns 43.875 percent of the royalty interest.
Using the math, 43.875% of $1.14 billion is $500 million, but UCLA says it will receive $520 million. (I contacted UCLA PR reps to clarify but so far haven't heard back). It may be that the $520 million includes both the $500 million and about $17 million (15% of $1.14 billion) for its chemistry department.

Still, this implies that Howard Hughes will receive about $70 million but the institute has not posted any mention on its media web page. And this leaves over $500 million to be split by the two lead inventors, other named inventors and possibly (as provided by UCLA policies) non-inventors who contributed towards development of the invention.

Thursday, February 4, 2016

Preparing the 21st century healthcare industry

Bob Curry, Ph.D.
As KGI prepares the strategic plan that will take it to its 25th anniversary, the members of the KGI community are being challenged to imagine where we need to be 2022 to continue to prepare cutting-edge graduates who will work across the healthcare value chain.

During a post-dinner exercise Wednesday night, we were encouraged to consider a map of imminent changes in healthcare suggested by Bob Curry, who is chairman of our Board of Trustees, a veteran VC, and now CEO of Perceptimed.

Based on this experience, Bob suggested four megatrends that KGI should consider:
  1. Growth in the science of diagnostic, prognostic, and monitoring tools will be explosive and be increasing paired with drug and procedure usage.
  2. Drugs will become every more customized to treat highly defined cohorts as characterized in the discussion point above. This will change both the nature of drug discovery and of clinical trail design.
  3. Healthcare will be delivered by a broader, integrated team of professionals than has been the traditional norm. Pharmacists and clinical diagnosticians will be teamed with physicians and nurses in staffing the healthcare system.
  4. Hospital systems and health insurance companies, as we currently recognize them, will disappear and will be replaced by 50-100 integrated care organizations to cover the entire U.S. (e.g. 50-100 Kaiser-like organizations).
The first two points appealed to our scientists in the room, who (since our 1997 founding) have been thinking about genomic and personalized medicine. The third point relates to those interested in clinical care delivery, particularly our pharmacy school which trains its PharmD students to work in teams with MDs and RNs. The final point ties to the business side — our group within KGI — and the changes brought by ACA (Obamacare), both pushed by the strong patient incentives for adverse selection and pulled by incentives for Affordable Care Organizations.

Based on this, the 70+ trustees, faculty, staff and students at nine tables generated a series of ideas. From their ideas — and my own observations — I see four important trends:
  • The importance of big data and data analytics. This is not just for analyzing genomic data for personalized medicine, but for patterns of clinical and other bioinformatic data for efficacy, drug-drug interactions, and other healthcare outcomes.
  • New types of healthcare providers and business models for funding them. 
  • Increasing importance of healthcare economics. Whether it’s HMOs, ACOs, capitation models, bundled pricing, or other approaches, we are moving away from a fee-for-service and dollars-per-pill model toward outcomes-based compensation.
  • New regulatory approaches to deal with these changes.
Some of these trends were building and accelerating over the past two decades. (I've been with one HMO for 30 years). Others were accelerated by the ACA. Still others (the destruction of insurance companies) were not among the announced goals of the ACA, but may be its inevitable outcome.

Friday, November 20, 2015

Barriers to Innovation in US Healthcare

This week I’m at #WOIC2015 (World Open Innovation Conference 2015) in Santa Clara. I am program chair for this the second annual conference, which was organized by the Garwood Center at the Haas School of Business at UC Berkeley.

This morning, the opening panel for the second day was on open innovation in healthcare within (and across) ecosystems.
Pramod John, William Bonfield, Amir Rubin, Sangita Reddy
The session opened with a keynote by Reddy, daughter of cardiologist Prathap Reddy (who founded the chain in 1983). The company has 57 hospitals (7 with US accreditation), 2,500 pharmacies and numerous doctors and outpatient clinics.

She talked about how the company used frugal innovation to provide solutions both to India’s middle class and poor. For example, the company has created a national telemedicine program that has touched 36 million patients. To support that, it’s created a device for remote testing of vital signs, and is working on a device for diagnosing malaria (and other parasite) infections.

Perhaps the best example was open heart surgery. Apollo has done 150,000 surgeries with a 99.3% success rate — and an average cost of $4,000. Yes, compared to the US it has lower labor and pharma costs. However, the clinicians pioneered a process (and clamp) to allow 89% of the surgeries be done as beating heart surgery — saving the machine that oxygenates the heart, the process (and risk) of starting/stopping the heart, and the longer recovery period.

Several speakers noted that the U.S. is still the gold standard for the newest, most advanced, most complicated cases. As Reddy said, “Advanced healthcare in the United States guides advanced healthcare in the rest of the world,” and speakers expressed concern about any changes that would eliminate the spillover values that provides for global medicine.

The U.S. problems are both in the incentives and the inefficiencies (including rent-seeking) in the current system. (Rubin faces specific challenges of real estate and labor costs in Silicon Valley, with nurses drawing $160k/year vs. $50k nationwide). As Bonfield remarked, the success of the system keeping people alive longer means there are more patients living with (expensive) chronic conditions.

On the inefficiencies, John argues that the biggest opportunity is in pharmaceutical distribution. Drug prices are rising while medical procedures are relatively flat. He estimated that 20% of the $400b annual pharma costs are wasted in the distribution channels, through pharmacy benefit manager (PBMs) and retailers. Using the website GoodRx.com, he offered examples of the same (generic) statin drug having more than a 5x range of retail pricing in a specific local neighborhood.

Not surprisingly, John has a tool to facilitate search for lower drug prices. Although it could be used by the uninsured or those with high deductible plans, their target is medium-sized firms that self-insure their pharmaceutical expenditures.
As the person who (successfully) pushed for KGI’s healthcare economics and drug pricing classes, I asked about the incentives. It’s great if I can save drug costs, but if it doesn’t budget my monthly premiums, I’m not going to bother. As one speaker notes, Singapore spends less than almost any developed country on healthcare (4.6% in 2013) through high deductibles and incentives for consumers to reduce their own costs (as you would on any other good).

Still, it was good to see suggestions of bottom-up innovation that have a real chance to bend the cost curve in a way that top down mandates cannot.

Friday, June 12, 2015

Accelerating life science startups

On Thursday, I attended the Indie Bio SF Demo Day, a coming out party for 12 companies in its San Francisco accelerator.

SOSVentures, the sponsor of Indie Bio (which has a second site in Cork, Ireland) believes this is one of the first life science accelerators.

As with a life science incubator, the accelerator requires provide startups with shared wet lab space. However, following the accelerator model pioneered by Y Combinator, the accelerator provides mentorship with a fixed term of residency.

The 12 companies applied in January, joined in February and marked their coming out Thursday with a pitch and exhibition to the press and potential investors. The 12 companies are:
  • ABioBot (Raghu Machiraju, CEO): lower cost, higher reliability lab automation
  • Affinity Wulfrun (Anil Bagha, CEO): improved column for manufacturing monoclonal antibodies
  • Arcturus BioCloud (Jamie Sotomayo, CEO): cloud hosted recombinant DNA experiments
  • ArkReactor aka Sensa.io: inexpensive bioreactors
  • BioLoom (Jennifer Kaehms, CEO): biomaterials for skin repair
  • Blue Turtle Bio (Adham Aljahmi, CEO): oral administration of enzyme replacement therapy
  • Clara Foods (Arturo Elizondo, CEO): synthetic egg whites
  • Extem Bioscience (Mardonn Chua, CEO): high throughput stem cell production
  • Orphidia (Aron Rachamim, CEO): point-of-care lab-quality assays from a single drop of blood
  • Pembient (Matthew Markus, CEO): synthetic rhino horn to supplant poaching
  • Ranomics (Cathy Tie, CEO): genomic database for oncology diagnostics
  • ZymoChem (Harshal Chokhawala, CEO): higher yield synthetic petrochemicals
TechCrunch profiled 11 of the 12 companies in their report of the demo event, while three of the companies were profiled when Co.Exist toured the IndieBio lab in April.

Talking to the entrepreneurs, all were indoctrinated in the "lean startup" philosophy. At least one of the companies has already done a “pivot.”

Consistent with that, each of the firms was trying to get to market with the minimum possible cash. Several of the companies have revenues already, and at least two hope to be cash flow positive within the next year. Several of these are tools companies — a business model that is quick to cash flow positive — while the one therapeutics company is targeting orphan diseases which offer a quicker and less expensive regulatory pathway.