Your first recommendation, i.e., eliminate GMPs for Phase 1 trial drugs, is based on a misunderstanding. In fact, FDA issued a final rule in July 2008 exempting Phase 1 drugs from the full GMP regulations and issued guidance that same month on what GMPs such drugs should follow. That guidance has about 8 pages of recommendations, and includes two sentences on stability:
"We recommend initiation of a stability study using representative samples of the phase 1
investigational drug to monitor the stability and quality of the phase 1 investigational drug during the clinical trial (i.e., date of manufacture through date of last administration)."
If a drug degrades significantly during the clinical study period it may cause direct harm and perhaps lead to a false conclusion about the drug's safety.
FDA rarely inspects clinical trial production and has always preferred a light regulatory touch at this stage of drug development. FDA reform is needed, but would prefer it be based on the truth and data.
Great post, I made a similar point below. It's worth highlighting that in my professional experience, FDA staff have generally been highly competent and scientifically informed, particularly compared to other regulators abroad. It's fair to argue that the FDA should be willing to tolerate more risk. Still, the narrative in this post of "look at how dumb FDA regulators are" does not match my professional experience working with the FDA.
As a gene therapy manufacturing professional, I see some errors in the GMP section, which makes me concerned for the general accuracy of this post.
> Or take AAV manufacturing—the FDA requires both a potency assay and an infectivity assay, even though potency necessarily reflects infectivity.
This is misleading. For AAVs, the FDA does not require true infectivity testing via TCID50 or comparable assay. The expectation is that you measure the empty/full ratio (i.e. the proportion of viral particles which contain the gene to be delivered). This test ensures patients aren't exposed to excess viral particles which is a safety risk; this safety risk is not addressed in potency testing.
> for example, the FDA often requires three months of stability testing for a drug patients will receive after two weeks.
This is again misleading. Autologous cell products are the one small subset of products where patients will receive a drug within two weeks of manufacturing. I'm not aware of a 3-month stability requirement for such products. For the vast majority of drugs there will be at least a few months between the manufacturing date and the date the drug is administered to the patient. Even for those products there is not a hard requirement to have 3 months of stability data; it will be up to the discretion of individual reviewers. All that is required by law is that you will demonstrate that the drug is stable until the last patient in the trial receives the drug. In the gene therapy guidance, they even explicitly say that you don't need to have stability data to support the full clinical trial when you file the initial IND, with the expectation being that you will submit an IND amendment later on in development to justify stability for the entire trial.
My dream FDA reform would be a rule prohibiting the exclusion of persons from clinical trials on the basis of race or biological sex, unless such exclusion was shown to be medically necessary.
Let’s just throw out the FDA and let seriously ill and terminal patients make their own choices. This is just more of the same: good or better regulation is still not liberty. Why can’t we make our own decisions about what to put into our bodies?!
Lab grown meat is no longer science fiction. Are we moving too fast?
While it promises sustainability and animal welfare, questions around health, bioavailability, and long term impact remain. In tomorrow’s post, I’ll break down what the research really says. No hype, no fear, just facts and nourishment forward insight.
We’ll explore studies from Frontiers in Nutrition, UC Davis, and Nature Food to understand whether lab grown meat truly nourishes us or simply looks like it does.
Kudos for covering so much ground in a relatively short piece. Lots to agree with and I appreciated the piece on expediting regulatory approval. Having worked with reg folks, it’s fair to say that they are not the most agile group of people so essentially revolutionising the reg function will be ..interesting. Also, fixing manufacturing is a major programme in and of itself. You can’t do this without changing the corp tax rate to kill transfer pricing. I don’t see generic manufacturing returning to the US in the absence of real manufacturing innovation to lower COGS. Finally, the problems wrt rare disease trials are real enough but I don’t see how you demonstrate stat significance with very small “n”. So in that case, you simply wave the product through (there are precedents here) and let the insurers make the call.
This is excellent. Implementing many of these changes would help the consumer. We spend so much time in pharma trying to ensure that we are compliant. It costs so much time and money and ultimately drives up the costs of drugs.
"What follows here is our prescription for the FDA (and in some cases HHS) — to accelerate and reduce the cost of new therapies [...]"
This is a great premise. What would be logical is to see some kind of deep analysis of what makes therapy prices in the US so high, and what slows down innovation (is it really slowed down, new drug approvals are close to all-time-high since 1962 Drug Amendments [1]) and by how much.
Instead, the assay jumps to conclusions that it's all (a) regulations, (b) regulations, and (c) regulations. Okay maybe not explicitly jumps to that conclusion, but all the prescriptions are basically "change regulations".
I doubt there are many countries in the world that have much simpler regulations of pharmaceuticals and because of that thrive in innovation and enjoy many more drugs than American patients. Comparison with China is misleading because the chart is about 'drugs being developed', while the rest of the assay is about 'drugs being approved'. The former is the problem of capital allocation (something FDA has only indirect impact on), the latter is related to regulation. However, no comparison of Chinese and American regulatory systems is given. Brief search led me to this recent paper [2] by Chinese authors that claims that "key milestones [in recent drug regulations] include the introduction of the Vaccine Management Law (2019) and stricter penalties under the Criminal Law Amendment (XI) (2020)." Doesn't sound like very innovation-stimulating things but what do I know - better analysis is always welcome.
Some listed arguments look weak. E.g., why need for preclinical reproductive safety if only non-pregnant people on contraceptives are allowed. Really, is that the best example of overregulation you could come up with? Irreversible infertility is one reason I could come up with while reading that paragraph without much mental effort. Again, by how much does this requirement slows down innovation? By few months? More likely less because multiple animal tests are done in parallel anyway - there's no regulation requiring them to be run sequentially (unlike clinical trials). How much cost does it adds, $10-100k to few $bn? How many similar little 'unnecessary' requirements are there - from the given example doesn't feel like too many.
What is presented as unequal treatment of US and foreign manufacturers looks to me more like a problem of FDA being understaffed. Same assay argues for patients' right to try more medicines but presents relaxed approval criteria during acute drug shortage as a bad thing. It's important to understand that these _are_ fundamental tradeoffs.
I'll spare time addressing other arguments because anyway to me it looks like the key problem - why drugs are so expensive in US - is not addressed at all. So not to be on all negative, here's my chain of thoughts.
Pharmaceuticals are known to be only ~10% of healthcare expenditure in the US. However, the way they are priced is important. Sure, R&D and approval process are expensive, but the pricing of drugs follows different - market - logic. The drugs are priced based on the currently available _alternative_ method of treatment. So if there's a cheap alternative drug - no way one can charge $10000 per pill and sell much. But for new truly innovative first-in-class drugs for uncurable diseases the alternative care is basically cost of pharmaceutical, hospital, and outpatient palliative care for the rest of patient's life - i.e. really expensive. That's why Tecfidera is priced at $2k per 14 capsules and generated billions in sales per year while the active ingredient is an industrial chemical produced by metric tons for pennies per kg. The whole development of the Tecfidera drug is estimated to be well below $1bn, so it did pay off ~100x any associated R&D costs, moreover it was basically paying multiple of Biogen's _entire_ annual R&D expenditure until 2020, and more than a half of it afterwards. And that's why Martin Shkreli could get away with his 5000% price hike of Daraprim. This IS the part that CAN be addressed with better regulations and is implicitly covered in the assay, but I digress.
The next step in the chain of thoughts is what makes _alternative_ care so expensive in the US compared to the rest of the world? Cost of hospital and outpatient care in the US is notoriously high. So driving it down should automatically drive down drug prices. What contributes to high hospital and outpatient costs that is very different from the rest of the world? The answer is of course very complex but the major thing that comes to my mind is relatively high doctor's salary (few fold higher than in Europe, excluding Switzerland [3]). Which in turn is based on the high expenses of medical training due to generally expensive American education and one of the longest durations of it in the US. Deregulation in these areas - aimed to increase domestic doctors supply - would have the most significant impact on how expensive drugs are. But that's far away from FDA's scope.
All in all the assay seems more ideologically than rationally charged. And like many initiatives championed by current administration it suffers from internal contradictions.
I was very fortunate to have worked with the founders of Genentech and Amgen at a startup biotech in Seattle, ICOS. cGMP standards are burdensome, but they are there for a reason: to ensure the product is made precisely the same each time. I know their importance, but they do impede innovation. However, when you are not in the discovery phase, they are necessary. It takes a certain kind of person to do that work. It does not require innovation, optimization, or a lot of thought. That said, it is in making mistakes that discoveries can be made, e.g., a 30X improvement in production.
At Pfizer, I saw the Sales and Marketing side, which is an interesting business side. I know RFK Jr. wants to kill marketing on TV. However, this is how many people suffering from a condition find out about therapies. Both Cialis and Viagra were initially developed for the vast cardiovascular market. In the human safety studies, past the animal studies, non-placebo patients were coming back for more. This is usually a sign of addiction. Finding out the real reason was difficult. However, as soon as it was discovered that they addressed erectile dysfunction (ED), that indication was filed. The big challenge that the drug companies faced was that people were embarrassed to ask for the drug. So, TV ads helped make it easier, and educating doctors on conditions often correlated with ED helped them begin the discussion with their patients. And NASCAR helped too.
The funny part about Viagra and Cialis was that they were approved for the original condition, i.e., as a CV drug. Compounding Pharmacies discovered this by accident and just changed the dosage by a milligram or two higher than the ED drug. This allowed them to sell the drugs at a generic price. Patents are important, but they expire rapidly. Workarounds do happen.
These are all great concepts—one more thing to consider: Updating the GRAS designation. By allowing thousands of chemicals to be grandfathered in under this status—and leaving it to industry to meet only minimal standards—the safety and health of Americans have suffered tremendously. This issue not only poses risks to soil, water, and broader environmental integrity, but it also permits these chemicals in cosmetics, foods, drugs, and devices. Regulatory oversight in this area has been far too lax. The Covid mRNA vaccine’s components are perfect examples of this! The PEGYLATED liquid nano particles used to support the messenger RNA work, only studied in a cursory fashion. They were proven to have cleared the body by only studying the water component of plasma, even though they are fat soluble chemicals. Their inflammatory nature was concealed until post market. And then safety data was only assessed, for the most part, by passive reporting using vigilance systems that are well known for being far less than adequate!
How can these be all great concept but all the rest of the comment, which completely contradicts the spirit of the assay? Aren't these fundamental tradeoffs: one can either have faster innovation due to deregulation OR slower but safer innovation.
These are all important points. Dr. Makary is a courageous leader who has never been afraid to take on big projects or to challenge academics. I’m convinced he’ll embrace many of the specific tasks you’ve recommended here. On the subject of clinical trials, the existing system is archaic and Balkanized. For many in care delivery, clinicaltrials.gov is hard to navigate. Also, academic medical centers limit access to the most efficacious trials in favor of their own. Health system RVUs also limit the incentive for doctors to invest themselves in the administrative burden of onboarding a patient into trials. Care delivery innovators nationally are creating ambulatory infrastructure and service models to remove the barriers to trials especially outside of AMCs and in rural areas. Trials are critical for patients who have failed standard of care and also for large self funded employers who bear the cost of catastrophic risk which is removed when trials are initiated. Great FDA leadership can inspire more and better innovation, easier access and broader and more efficient enrollment into the most efficacious trials. In the meantime, destroying barriers to entry for innovators should be the first principle.
In Free to Choose, Milton Friedman explains not only how to make FDA great, but also why we don’t need it at all.
Do we have what it takes to get rid of FDA?
Your first recommendation, i.e., eliminate GMPs for Phase 1 trial drugs, is based on a misunderstanding. In fact, FDA issued a final rule in July 2008 exempting Phase 1 drugs from the full GMP regulations and issued guidance that same month on what GMPs such drugs should follow. That guidance has about 8 pages of recommendations, and includes two sentences on stability:
"We recommend initiation of a stability study using representative samples of the phase 1
investigational drug to monitor the stability and quality of the phase 1 investigational drug during the clinical trial (i.e., date of manufacture through date of last administration)."
If a drug degrades significantly during the clinical study period it may cause direct harm and perhaps lead to a false conclusion about the drug's safety.
FDA rarely inspects clinical trial production and has always preferred a light regulatory touch at this stage of drug development. FDA reform is needed, but would prefer it be based on the truth and data.
Great post, I made a similar point below. It's worth highlighting that in my professional experience, FDA staff have generally been highly competent and scientifically informed, particularly compared to other regulators abroad. It's fair to argue that the FDA should be willing to tolerate more risk. Still, the narrative in this post of "look at how dumb FDA regulators are" does not match my professional experience working with the FDA.
As a gene therapy manufacturing professional, I see some errors in the GMP section, which makes me concerned for the general accuracy of this post.
> Or take AAV manufacturing—the FDA requires both a potency assay and an infectivity assay, even though potency necessarily reflects infectivity.
This is misleading. For AAVs, the FDA does not require true infectivity testing via TCID50 or comparable assay. The expectation is that you measure the empty/full ratio (i.e. the proportion of viral particles which contain the gene to be delivered). This test ensures patients aren't exposed to excess viral particles which is a safety risk; this safety risk is not addressed in potency testing.
> for example, the FDA often requires three months of stability testing for a drug patients will receive after two weeks.
This is again misleading. Autologous cell products are the one small subset of products where patients will receive a drug within two weeks of manufacturing. I'm not aware of a 3-month stability requirement for such products. For the vast majority of drugs there will be at least a few months between the manufacturing date and the date the drug is administered to the patient. Even for those products there is not a hard requirement to have 3 months of stability data; it will be up to the discretion of individual reviewers. All that is required by law is that you will demonstrate that the drug is stable until the last patient in the trial receives the drug. In the gene therapy guidance, they even explicitly say that you don't need to have stability data to support the full clinical trial when you file the initial IND, with the expectation being that you will submit an IND amendment later on in development to justify stability for the entire trial.
Certainly a better and more productive use of Trump admin time and resources than tariffs.
My dream FDA reform would be a rule prohibiting the exclusion of persons from clinical trials on the basis of race or biological sex, unless such exclusion was shown to be medically necessary.
Let’s just throw out the FDA and let seriously ill and terminal patients make their own choices. This is just more of the same: good or better regulation is still not liberty. Why can’t we make our own decisions about what to put into our bodies?!
Coming tomorrow on The Nourished Standard
Lab grown meat is no longer science fiction. Are we moving too fast?
While it promises sustainability and animal welfare, questions around health, bioavailability, and long term impact remain. In tomorrow’s post, I’ll break down what the research really says. No hype, no fear, just facts and nourishment forward insight.
We’ll explore studies from Frontiers in Nutrition, UC Davis, and Nature Food to understand whether lab grown meat truly nourishes us or simply looks like it does.
Let’s keep this community thoughtful and curious.
Kudos for covering so much ground in a relatively short piece. Lots to agree with and I appreciated the piece on expediting regulatory approval. Having worked with reg folks, it’s fair to say that they are not the most agile group of people so essentially revolutionising the reg function will be ..interesting. Also, fixing manufacturing is a major programme in and of itself. You can’t do this without changing the corp tax rate to kill transfer pricing. I don’t see generic manufacturing returning to the US in the absence of real manufacturing innovation to lower COGS. Finally, the problems wrt rare disease trials are real enough but I don’t see how you demonstrate stat significance with very small “n”. So in that case, you simply wave the product through (there are precedents here) and let the insurers make the call.
Potential correction: I don't think the NTD PRV program failed to be reauthorized. Just the medical countermeasures one.
For more on why expanding the PRV program to include addiction indications is so high-impact, see this section of the Innovation Agenda for Addiction:
https://recursiveadaptation.com/i/152890853/building-a-robust-market-in-addiction-medicine-development
This is excellent. Implementing many of these changes would help the consumer. We spend so much time in pharma trying to ensure that we are compliant. It costs so much time and money and ultimately drives up the costs of drugs.
Missing “is” in “It déjà vu: we have…” Correct then delete this comment.
"What follows here is our prescription for the FDA (and in some cases HHS) — to accelerate and reduce the cost of new therapies [...]"
This is a great premise. What would be logical is to see some kind of deep analysis of what makes therapy prices in the US so high, and what slows down innovation (is it really slowed down, new drug approvals are close to all-time-high since 1962 Drug Amendments [1]) and by how much.
Instead, the assay jumps to conclusions that it's all (a) regulations, (b) regulations, and (c) regulations. Okay maybe not explicitly jumps to that conclusion, but all the prescriptions are basically "change regulations".
I doubt there are many countries in the world that have much simpler regulations of pharmaceuticals and because of that thrive in innovation and enjoy many more drugs than American patients. Comparison with China is misleading because the chart is about 'drugs being developed', while the rest of the assay is about 'drugs being approved'. The former is the problem of capital allocation (something FDA has only indirect impact on), the latter is related to regulation. However, no comparison of Chinese and American regulatory systems is given. Brief search led me to this recent paper [2] by Chinese authors that claims that "key milestones [in recent drug regulations] include the introduction of the Vaccine Management Law (2019) and stricter penalties under the Criminal Law Amendment (XI) (2020)." Doesn't sound like very innovation-stimulating things but what do I know - better analysis is always welcome.
Some listed arguments look weak. E.g., why need for preclinical reproductive safety if only non-pregnant people on contraceptives are allowed. Really, is that the best example of overregulation you could come up with? Irreversible infertility is one reason I could come up with while reading that paragraph without much mental effort. Again, by how much does this requirement slows down innovation? By few months? More likely less because multiple animal tests are done in parallel anyway - there's no regulation requiring them to be run sequentially (unlike clinical trials). How much cost does it adds, $10-100k to few $bn? How many similar little 'unnecessary' requirements are there - from the given example doesn't feel like too many.
What is presented as unequal treatment of US and foreign manufacturers looks to me more like a problem of FDA being understaffed. Same assay argues for patients' right to try more medicines but presents relaxed approval criteria during acute drug shortage as a bad thing. It's important to understand that these _are_ fundamental tradeoffs.
I'll spare time addressing other arguments because anyway to me it looks like the key problem - why drugs are so expensive in US - is not addressed at all. So not to be on all negative, here's my chain of thoughts.
Pharmaceuticals are known to be only ~10% of healthcare expenditure in the US. However, the way they are priced is important. Sure, R&D and approval process are expensive, but the pricing of drugs follows different - market - logic. The drugs are priced based on the currently available _alternative_ method of treatment. So if there's a cheap alternative drug - no way one can charge $10000 per pill and sell much. But for new truly innovative first-in-class drugs for uncurable diseases the alternative care is basically cost of pharmaceutical, hospital, and outpatient palliative care for the rest of patient's life - i.e. really expensive. That's why Tecfidera is priced at $2k per 14 capsules and generated billions in sales per year while the active ingredient is an industrial chemical produced by metric tons for pennies per kg. The whole development of the Tecfidera drug is estimated to be well below $1bn, so it did pay off ~100x any associated R&D costs, moreover it was basically paying multiple of Biogen's _entire_ annual R&D expenditure until 2020, and more than a half of it afterwards. And that's why Martin Shkreli could get away with his 5000% price hike of Daraprim. This IS the part that CAN be addressed with better regulations and is implicitly covered in the assay, but I digress.
The next step in the chain of thoughts is what makes _alternative_ care so expensive in the US compared to the rest of the world? Cost of hospital and outpatient care in the US is notoriously high. So driving it down should automatically drive down drug prices. What contributes to high hospital and outpatient costs that is very different from the rest of the world? The answer is of course very complex but the major thing that comes to my mind is relatively high doctor's salary (few fold higher than in Europe, excluding Switzerland [3]). Which in turn is based on the high expenses of medical training due to generally expensive American education and one of the longest durations of it in the US. Deregulation in these areas - aimed to increase domestic doctors supply - would have the most significant impact on how expensive drugs are. But that's far away from FDA's scope.
All in all the assay seems more ideologically than rationally charged. And like many initiatives championed by current administration it suffers from internal contradictions.
[1] https://media.nature.com/lw767/magazine-assets/d41573-025-00001-5/d41573-025-00001-5_50411856.jpg
[2] https://www.mdpi.com/2227-9032/13/6/588
[3] maybe it's worth it to compare US regulatory system with Swiss?
I was very fortunate to have worked with the founders of Genentech and Amgen at a startup biotech in Seattle, ICOS. cGMP standards are burdensome, but they are there for a reason: to ensure the product is made precisely the same each time. I know their importance, but they do impede innovation. However, when you are not in the discovery phase, they are necessary. It takes a certain kind of person to do that work. It does not require innovation, optimization, or a lot of thought. That said, it is in making mistakes that discoveries can be made, e.g., a 30X improvement in production.
At Pfizer, I saw the Sales and Marketing side, which is an interesting business side. I know RFK Jr. wants to kill marketing on TV. However, this is how many people suffering from a condition find out about therapies. Both Cialis and Viagra were initially developed for the vast cardiovascular market. In the human safety studies, past the animal studies, non-placebo patients were coming back for more. This is usually a sign of addiction. Finding out the real reason was difficult. However, as soon as it was discovered that they addressed erectile dysfunction (ED), that indication was filed. The big challenge that the drug companies faced was that people were embarrassed to ask for the drug. So, TV ads helped make it easier, and educating doctors on conditions often correlated with ED helped them begin the discussion with their patients. And NASCAR helped too.
The funny part about Viagra and Cialis was that they were approved for the original condition, i.e., as a CV drug. Compounding Pharmacies discovered this by accident and just changed the dosage by a milligram or two higher than the ED drug. This allowed them to sell the drugs at a generic price. Patents are important, but they expire rapidly. Workarounds do happen.
These are all great concepts—one more thing to consider: Updating the GRAS designation. By allowing thousands of chemicals to be grandfathered in under this status—and leaving it to industry to meet only minimal standards—the safety and health of Americans have suffered tremendously. This issue not only poses risks to soil, water, and broader environmental integrity, but it also permits these chemicals in cosmetics, foods, drugs, and devices. Regulatory oversight in this area has been far too lax. The Covid mRNA vaccine’s components are perfect examples of this! The PEGYLATED liquid nano particles used to support the messenger RNA work, only studied in a cursory fashion. They were proven to have cleared the body by only studying the water component of plasma, even though they are fat soluble chemicals. Their inflammatory nature was concealed until post market. And then safety data was only assessed, for the most part, by passive reporting using vigilance systems that are well known for being far less than adequate!
How can these be all great concept but all the rest of the comment, which completely contradicts the spirit of the assay? Aren't these fundamental tradeoffs: one can either have faster innovation due to deregulation OR slower but safer innovation.
These are all important points. Dr. Makary is a courageous leader who has never been afraid to take on big projects or to challenge academics. I’m convinced he’ll embrace many of the specific tasks you’ve recommended here. On the subject of clinical trials, the existing system is archaic and Balkanized. For many in care delivery, clinicaltrials.gov is hard to navigate. Also, academic medical centers limit access to the most efficacious trials in favor of their own. Health system RVUs also limit the incentive for doctors to invest themselves in the administrative burden of onboarding a patient into trials. Care delivery innovators nationally are creating ambulatory infrastructure and service models to remove the barriers to trials especially outside of AMCs and in rural areas. Trials are critical for patients who have failed standard of care and also for large self funded employers who bear the cost of catastrophic risk which is removed when trials are initiated. Great FDA leadership can inspire more and better innovation, easier access and broader and more efficient enrollment into the most efficacious trials. In the meantime, destroying barriers to entry for innovators should be the first principle.