2/18/2019

921-922 days left – 30 months remaining

Biopharma/Investing

The next deep dive will be on REGN.

Here are 10 biotech surprises for the next few years. They are in no particular order.

  1. Microbiome comes up empty.
    There’s not enough here to create true clinical applications. Some of the findings are real but so often the solutions will be cheap antibiotics, leaving VCs (generally bad investors) holding the bag.
  2. Biosimilars never get scale and stink as a business.
    Despite Sandoz having $1 billion in biologics revenue, few companies purveying biosimilars have hit even 1/10th of this milestone. The key problem is without mandatory switching/substitutability, only new patients will get biosimilars. With brand discounting and cautious physicians not wanting to mess with what works (physician rule #1), taking some share of just new patients when almost all of these drugs are life-long chronic therapy is not exciting. Furthermore, by the time a drug becomes biosimilar eligible, there are generally creeping alternatives (exception: Neulasta).
  3. Immunooncology has no second act.
    With PD-1 being the first act (we’ll talk about CARTs in a moment), I don’t think any new “IO” drugs will be meaningfully efficacious. The situation will be akin to 2006-2008’s hunt for angiogenesis inhibitors after Avastin’s success. Turns out you can only deplete blood vessel growth so much. I think it will turn out that you can only prime the immune system so much before bad things happen. I think oncology will be a great area to invest with respect to mutational drivers, tumor microenvironment, metastases drivers, ADCs, and other modalities. IO is finished. RIP GITR, IL-2, TIGIT, OX40, IDO, etc.
  4. CART not useful, off-the-shelf therapies dominate.
    I only wanted to include shocking predictions in this list. With the 4th quarter of Kymriah sales coming in at $28 million, is this really shocking? There just isn’t enough juice in CART that makes it worth the squeeze.
  5. Don’t call major primary care categories being good business opportunities a comeback.
    The trend towards rare/expensive drugs has left major illnesses like osteoarthritis, obesity, pain, cardiovascular in general, etc. without much drug development love. I think the success of CGRP and potentially some new obesity and pain agents will remind folks that selling cheap drugs to millions of people is 1) a decent business plan and 2) avoids the high-price socialism we’re seeing on the fringes of the radical left.
  6. There will not be a new Alzheimer’s drug approved by the FDA for at least 10 more years.
    There is some emerging hope, however. Unfortunately I won’t say anything more on the subject.
  7. CAG/polyQ repeat disorders will not be treatable by the antisense modality.
    The one prediction I might be most wrong on, I believe we don’t have huntingtin, frataxin and other proteins just for them to cause lethal expansions. No, they are functional proteins after development and getting rid of the amyloid aggregates will not solve the problem.
  8. Despite the above nucleic acid therapeutics of all kinds will propser far more than anticipated.
    I think in the excitement of CRISPR and gene therapy, NA-based therapies may be overlooked as only a few companies have the know-how to deploy the chemistry required to be useful medicinally. As that technology passes into the public domain, more companies should employ these techniques.
  9. Orphan pricing will continue to grow unabated – $5 and $10 million per-patient-per-year therapies will be introduced.
    In a Moore’s law-type dynamic, PPPY has increased not because of some malefactors (and their epigones), but because society craves health more than any other good. Health spending since the 60s is on a power law that cannot be stopped by any legislation or NGO. It is not a bad thing.
  10. Artificial intelligence/molecular dynamics software will not improve drug discovery.
    These technologies have existed for decades and the advent of supercomputer clusters passing the baton to cloud computing does not significantly alter the status quo to generate any kind of revolution.

Papers I’ve Read

Universal Medicine Access through Lump-Sum Renumeration — Australia’s Approach to Hepatitis C. Moon & Erickson. NEJM 2019.
There’s a lot wrong with this paper. The authors predictably begin by crying: “high prices can restrict access to medicines in rich and poor countries alike”. Nevermind the determine of what is “high”–virtually any price is high, isn’t it? Australia’s “Netflix” model is put forth as a potential sword against drug pricing. Laughable. What doesn’t dawn on the authors, two (too-) inexperienced academics, is that Australia is 2-4% (being generous) of drug spend. Companies like Gilead don’t focus on Australia and couldn’t care less about maximizing revenue in this territory. The paper humorously ignores the net pricing of DAA in the US, instead assuming list price. We should all be so lucky to live in the world these authors do. It’s always sad to read editorials like this in well-meaning and ordinarily great journals like NEJM. Feelings. Reason. Who should win?

13 thoughts on “2/18/2019”

  1. Thanks for the post. Hope all is well.

    Re: #10 — I would agree that the shift in underlying infrastructure (from on-premise to cloud) isn’t transformative. But to my understanding, that change in infra isn’t the focus; rather, the AI story of today seems to be more about the emergence of *deep learning* as the most dominant/promising technique (out of the many subdisciplines that were, until recently, all fighting underneath the same “AI” umbrella). That said, I would be interested to learn more about the historical role neural nets have played in drug discovery up until now, and how the most recent advancements in the study can be applied going forward. Any thoughts?

    1. Dear Martin,
      Dear Trashy,

      I would like to chip in here and ask for a further breakdown of the reasons for point #10.

      To me (apart from the data quality problem), it seems that the potential of “AI” (be it linear statistics, deep learning, or other concepts) usage is still in the exploration phase.

      Cheers,

      M.

  2. I think microdialysis is a requirement for expanding our knowledge about “tagged” MHCs & how to use our immune system effectively. I’m thinking something akin to surface chemistry & electrochemistry. I don’t think it’s a lost cause, just needs outside thinking applied.

  3. What are your thoughts on Actinogen and their 11β-HSD1 inhibitor UE2343 / Xanamem considering the upcoming phase II results?

  4. You most definately did recommend ABEO. Your post from 8/30/18: “Abeona (ABEO), Axovant (AXON), Audentes (BOLD), BioMarin (BMRN), Oxford, PTC (PTCT), Spark (ONCE), UniQure (QURE), Rocket (RCKT), Sangamo (SGMO), RGNX, Sarepta (SRPT), Solid (SLDB), Ultragenyx (RARE), Voyager (VYGR) are all players. My guess is an equal-weighted or capitalization-weighted index of the above would outperform pretty much any other index.”

Leave a Reply

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

This site uses Akismet to reduce spam. Learn how your comment data is processed.