Hacker News Reader

Comment by Steuard | Hacker News Reader

It's an indictment of something that we've had this amazing new tool for vaccine production in existence for a decade but we never even tried to use it for that until COVID because "vaccines typically don’t make much money" and "vaccines just aren’t that exciting" scientifically: "it’s nice and easy," to quote Rossi from the article.

If it really does turn out that effective mRNA vaccines can be created in a straightforward way for a host of new diseases, that's ten years of deaths and misery that we as a society just allowed to happen unnecessarily. I don't know how to fix that: maybe the market-based incentive structures that work for acute disease treatments aren't a good fit for broad preventative public health measures like vaccines.

View on HN

Or focus on lowering costs. We learned in the pandemic how fundamentally broken the FDA is. There is no reason that things can't be faster that "operation warp speed" all the time.

Or re-focus the monetary models around either markets or patient outcomes. Right now, medicine is not a market, it's a strange game. The most money that can be made is finding a treatment that may provide a modest increased benefit over what is live, then sell it to medicare at "name your own price". This incentives high probability wins (which are typically things we already understand well).

By this model vaccines are sadly "too effective" on two counts:

- They are preventative. Americans are scared that regular people walk into hospitals and debt peons walk out. We are too sickly and scared for preventative medicine, and like our infrastructure which is never maintained to the point of being in "good working order".

- They are O(1) dose. The unity economics of once-and-done is bad. Better to sell something which makes a customer for life.

It's really depressing.

> There is no reason that things can't be faster that "operation warp speed" all the time.

Because money. Operation warp speed started mass manufacturing at the same time as clinical trials. 70% of drugs fail the 3rd stage of trials, so that would be a hugely wasteful endeavor to parallelize this when failure is the norm.

> There is no reason that things can't be faster that "operation warp speed" all the time.

Safety? I'm not wild about new vaccines requiring emergency use authorization and indemnity for the manufacturer. It worked out fine for COVID, but this is not a model I would _want_ to replicate.

The authorization is eua not because of safety concerns but unknown long term efficacy > 1 year. Does this change your assessment?

We did not have this tool for a decade, it was still very much scientific work in progress, with a lot of practical problems to be solved. Those problems were only ironed out in 2017-2018.

You can absolutely kill or at least maim a new medical technology if you push it on the market too far and a wave of serious side effects hits you. As an example: death of Jesse Gelsinger [0] delayed genetic therapies by several years, possibly a decade or more. Scientists were afraid to touch the tools that killed a young man and produced a public backlash.

[0] https://en.wikipedia.org/wiki/Jesse_Gelsinger

This is not new at all. They tried to use mRNA vaccines for other diseases as well, but were never approved. This still holds true to this date.

edit: An example would be Dengvaxia.

It seems that Dengvaxia is a live attenuated vaccine, what is the relationship with mRNA vaccine?

Yeah, you're right. Better examples are:

Zika: https://clinicaltrials.gov/ct2/show/NCT04064905

Influenza: https://pubmed.ncbi.nlm.nih.gov/31079849/

Rabies: https://pubmed.ncbi.nlm.nih.gov/28754494/

I see two big problems.

The science funding climate, at least in the US (I don't really know what goes on elsewhere) is in a really sad state. Projects with a low probability of success are really hard to get funded but that is precisely what you need to come up with really novel things. Curiously, science funding bodies don't seem to think of things in a probabilistic way.

The second issue is the same idea, but applied to industry. There is an activation energy that you need to surpass in order to make any business profitable. If we are talking about building a business around a brand new and unproven (in a business sense) technology like mRNA, the barrier is _huge_. You need entities with very deep pockets to accept year after year of losses before you can start getting a viable business producing cutting-edge products.

In both cases I think intelligent use of government funds could be a big help. I am not big on socialism, but I have long held that this is one of the most important things a government can and should provide: subsidies for risky research and business endeavors, not to mention large-scale infrastructure projects.

We do fund basic research decently well, but not development. That's mostly in agreement of the above, but leads to a few differences.

In particular, if we start funding the both research & development, the public sector starts taking on all risk, so there is 0 reason to reward the private sector with IP ownership. Do a drug bounty, and then public domain the IP. I would say have the state hold the IP and license it out with cost controls, but I think having medicare actually negotiate will do that well enough.

I agree with your sentiment, but we also need to careful with hindsight bias (i.e. it's easy to pick the winners after the fact)! It's a tricky question!

Sure, but when urgency suddenly became necessary, mRNA technology was pursued quickly, which indicates that we already knew it was promising.

A good chunk of that is that people had finally gotten a delivery mechanism to work in the past few years. The "for a decade" is misleading, because a decade ago it was an interesting idea nobody had gotten to work in a way that was useful. 2020 it was something with good promise and product development under way that could be sped up massively with more money and acceptance of it being a gamble.

This feels like a post-hoc rationalization to me. Of course it was less ready, and then more ready. But exactly how much of a gamble was it?

The upside is immense even considering just existing diseases, but we are really bad at pricing things that improve the status quo vs prevent bad things.

The theory, assuming a delivery mechanism is found, is rock solid---"if not when"---and I therefore wouldn't call it a gamble.

On the other hand, I don't know enough about the biology to speak to the difficulty and uncertainty around delivery mechanisms. Can we get more info on that? I suspect it wasn't too uncertain, though "oh, we change this base pair a bit and the immune system doesn't care" does seem like a relatively unplanned discovery.

I believe the correct packaging mechanism was only conceived in 2018.

mRNA vaccines have been undergoing active research for decades, and in particular the lipid nanoparticle delivery mechanism used in both the Pfizer and Moderna vaccines was first used in 2018 in a (non-vaccine) drug called Onpattro[2]. Without this technology, Moderna in particular was having issues with excessive side effects as late as 2017[1].

As such, the Covid-19 vaccines from Pfizer/BioNTech and Moderna could not have been developed pre-2018, and - absent a deadly pandemic that requires taking extreme measures (such as using very large trial groups) to hasten the timeline - normally vaccines undergo development over a period of many years[3] to collect sufficient data to ensure effectiveness and to ensure that the long-term side effects are well understood. Moderna in particular was, prior to the Covid-19 pandemic, working on developing vaccines to target cancer using mRNA technology[4]. Had the pandemic not occurred, one would therefore not expect to have seen any approved therapies until the mid-to-late 2020s at the earliest (and probably later, given how often vaccines and other drugs fail in human trials).

In short, the issue was not that pharmaceutical companies were disregarding the technology, so much as simply the last remaining breakthrough needed to be made - and now that mRNA vaccines have proven themselves to be safe and effective, there'll be no shortage of effort to apply them to new diseases.

[1] - https://www.statnews.com/2017/01/10/moderna-trouble-mrna/ [2] - https://www.statnews.com/2020/12/01/how-nanotechnology-helps... [3] - https://www.historyofvaccines.org/content/articles/vaccine-d... [4] - https://www.statnews.com/2017/01/10/moderna-trouble-mrna/

I thought about a scheme where the years of patent protection in a particular field where there has been lack of progress is adjusted upwards, creating incentives for r&d in that particular space. The same obviously should also go into the other direction

I think patents are just a poor way to fund this stuff in general. Better to do drug bounties in which case the financing trickiness is only about who funds the human trials.

In more than medicine, there's a good argument that only monopolies can afford R&D (c.f. Bell Labs) and monopolies are also terrible, so I'm pretty down on leaving R&D to the private sector anyways. Remember we already publicly fund all the R, this just about the D.

Also, ownership, and IP ownership in general, far from being the natural order of things, is an extremely weird and hard to price financial instrument. I don't see the point of forcing it, and trying to make it work with e.g. adjustable patent lengths. That just feels like an epicycle that's very prone to regulatory capture.

Hehe try jumping in the https://en.wikipedia.org/wiki/Post-Keynesian_economics rabbit hole, where perhaps the simplified jumping off point is "growth follows demand".

That fact that we fund all this basic research, but then can hardly be bothered with the development part of "R&D" seems like a pretty strong anecdote in favor.

Reply on news.ycombinator.com