Tuesday, April 01, 2025

Federal Judge rules that FDA's attempt to regulate clinical laboratories exceeds its authority

Overview
On March 31, 2025, a federal court vacated the FDA's final rule attempting to regulate laboratory-developed tests (LDTs) as medical devices under the Federal Food, Drug, and Cosmetic Act (FDCA). The ruling is a significant victory for clinical laboratories across the country.

FDA regulates test kits (like home Covid test kits) as "medical devices" and has developed stringent requirements for their manufacture and sale. An LDT is a service, agreed between the patient's medical doctor and the medical doctor in the laboratory, to test the patient's specimen for various "analytes" (antibodies to viruses, hormones, viral particles, etc.). These tests are developed by the laboratory because manufactured kits are not available for the exact test the patient needs. The FDA had asserted a right to regulate LDTs as medical devices.  

The advantage of LDTs is that they can be tailored to fit these needs quickly and for far less cost than asking a kit manufacturer to create a specialized test and obtain FDA approval. LDTs are regulated by the Center For Medicare and Medicaid Services (CMS) and the Clinical Laboratory Improvements Amendments (CLIA) which require the lab to "establish performance characteristics" --prove that the test works as claimed-- and retain the proof for inspection by CMS. 

FDA's decision to regulate LDTs would have resulted in double regulation (CMS and FDA), slowed innovation and cost billions of dollars, which would ultimately be passed on the patient.

The rule would subject about 79,114 existing LDTs and 1,181 labs to FDA premarket reviews, post-market surveillance, and quality system requirements. "FDA acknowledged that the costs of the proposed rule would be significant. Specifically, it estimated that the cost of preparing and submitting [applications] for existing laboratory-developed tests would exceed $35 billion and could reach $113 billion. It also projected that, going forward, annual compliance costs for affected laboratories would range from $4 billion to $14 billion." (From the judgment linked below.)

On March 31, 2025, a federal court vacated the FDA's rule to regulate LDTs as medical devices, marking a significant legal setback for the agency. Judge Sean D. Jordan of the U.S. District Court for the Eastern District of Texas ruled that the FDA lacks statutory authority under the Federal Food, Drug, and Cosmetic Act (FDCA) and CLIA to oversee LDTs, which he characterized as diagnostic services rather than physical medical devices.

The court emphasized that Congress designed CLIA as the primary regulatory framework for clinical laboratories, delegating oversight to CMS rather than the FDA. LDTs were deemed "proprietary methodologies" rather than physical devices, placing them outside the FDA’s jurisdiction.


Of course, FDA may appeal the decision, or Congress may act and change the requirements by law. 


I know laboratories have already spent large amounts attempting to get ready for the May 2025 deadline. (I should have taken up the offer to act as a consultant...too late now.) My advice is to keep doing all the things the FDA has so far required. No harm, and possibly great good, will come from having properly documented complaints procedures and test development and change files beyond the original CLIA requirements. 


The judge's ruling in full.

Thursday, March 27, 2025

A Rant Concerning AI helpiness, and how terrible everything is nowadays

 Google search is useless. 

I know this will surprise you young whippersnappers, but back in the day, if you typed something in the Google search bar, it would look for and bring up documents that used those words.  This was a very useful feature. It didn't bring up any sponsored pages, ads or inane commentary supplied by the "AI." 

If you typed 

generals gathered masses just like witches

It would find the lyric sheet for the song "War Pigs" by Black Sabbath, and because those words all appear in the first few lines, that would be in the first page of results.

A couple of days ago, I typed 

the winter life is coming back

expecting the algorithm to name the song by the group Cream that contains the line. This is what I got.


"the winter life is coming back" AI Overview Learn more Given that it's March 22nd, 2025, and you're in San Juan Capistrano, California, it's unlikely that "the winter life" is returning, as you are in the midst of spring. San Juan Capistrano experiences mild winters and warm, dry summers.  Here's why: Time of Year: March 22nd marks the beginning of spring in the Northern Hemisphere, not the return of winter.  San Juan Capistrano Climate: San Juan Capistrano has a Mediterranean climate, characterized by mild, wet winters and warm, dry summers.
Screenshot of search and results

In other words, the AI just started arguing with me about whether the line in the query bar was true or not!

The song turns out to be, as I'm sure you're all eager to tell me, "Deserted Cities of the Heart" by Cream. 

I asked Google what it thought it was doing, and it replied, in part: 

does google do boolean search terms like it used to  Examples: "bacon eggs" (or bacon eggs) - finds pages containing both words  bacon OR eggs - finds pages containing either "bacon" or "eggs"  bacon -eggs - finds pages containing "bacon" but not "eggs"  "San Juan Capistrano" - finds pages with the exact phrase "San Juan Capistrano"
Screenshot


And this - Google's own AI's answer - is wrong, or at best highly misleading. 

"bacon eggs" written that way does not find pages containing both words. It finds pages where the words "bacon eggs" appear in exactly that order. (For example, "Serve it with bacon, eggs, and toast.")

To find pages where both bacon and eggs appear but not necessarily together in that order (for example, "people like to eat bacon with eggs"), you would have to type the words without quotation marks, as below:

bacon eggs

On Twitter, oh I'm sorry X, these days, I see a huge percentage of people answering a Tweet by asking Grok, Twitter's sorry X's AI, to explain what the Tweet meant, or if the Tweet was "true." Depending on the phrasing, Grok will say different things to different people. A whole host - millions of queries - using lakes' worth of water and thousands of dollars worth of electricity - of utterly useless questions and even more useless answers, many of which are probably taken as gospel by the question asker. 

No good can come of this. The incredible investment in AI cannot pay off for the companies building out capacity and pinning their business plans on some fairytale of return. A 2000-style dot com crash is imminent. The only lasting effect of AI will be dumbed-down people fighting in their bunkers for their interpretation of everything based on idiomatic peculiarities in their queries. One hopes (with little confidence) that the fighting will remain virtual and not spill into the real world.



Wednesday, March 26, 2025

San Juan Capistrano Swallows Day Parade 2025

I attended the San Juan Capistrano Swallows Day Parade, but didn't get around to seeing much because I was a little hoarse.

a small horse outside a white horse trailer

The 2025 San Juan Capistrano Swallows Day Parade was held on Saturday, March 22nd. This event is part of the broader Fiesta de las Golondrinas, which celebrates the return of the swallows to the San Juan Capistrano Mission, a tradition that coincides with St. Joseph's Day on March 19th.

The parade is a significant event in San Juan Capistrano, highlighting the city's rich equestrian heritage and featuring equestrian teams, dancers, and marching bands. The theme for the 2025 parade was "Lassos and Lace." 

The tradition of celebrating the swallows' return began in the 1920s, initiated by Father O'Sullivan, who encouraged the swallows to nest at the Mission since their nests were being destroyed elsewhere in the town. Over time, this event has grown into a major cultural and tourist attraction for San Juan Capistrano.

A person in a bull costume drives a covered wagon
One of the paraders outside the Mission


Sunday, March 23, 2025

Oathbreaker (short story) Published in Black Cat Weekly today

Cover illustration, dragon and space suited woman, title of magazine Black Cat Weekly and list of contributors including lyle hopwood
Cover, Black Cat Weekly #186

I'm thrilled to announce that my short story "Oathbreaker" appears in Black Cat Weekly #186, out today.

It's the adventure tale of a Welsh professor whose trip to the coast to teach a course on seaweed is hijacked by ghosts who need help rescuing a woman enslaved to a cursed Viking raider a thousand years ago.

Black Cat Weekly is an action-packed magazine - there's an entire novel in this one, along with the usual short stories and features, for just $2.99 (ebook).


 

Wednesday, March 05, 2025

Bioscience Ethics at 50: Asilomar and the beginning of genetic engineering

In 1973 the Gordon Conference on Nucleic Acids took place at New Hampton, Hampshire, where advances in recombinant DNA technology were discussed. The scientists had concerns about the ramifications of the new technology. Was there a possibility of harmful organisms escaping the laboratory?

One researcher, Paul Berg (who went on to a very distinguished career and died only recently) was working with SV40, a cancer-producing simian virus (as an aside, SV40 had been accidentally injected into many polio vaccine recipients in the 1950-1960s).  Berg pioneered using restriction enzymes to cut DNA loops and then using other enzymes to rejoin them, creating new combinations of genetic material. Others at the same time were working on similar “recombinant” technology.

Here’s a video of Berg discussing his work and his belief that a “moratorium” of this technology should be called. Partial transcript of the video:

The point was that we were going to be propagating SV40 genes in E. coli. The concern was that E. coli is a natural inhabitant in humans, and the worry was whether we could inadvertently infect humans with the E. coli we were propagating, potentially leading to cancer. This concern sparked intense debate. Depending on where you stood on the issue, you were either vociferous about it or dismissive, saying, "Forget about it. It's not serious."

After many conversations, I became convinced that I could not dismiss the risk and make it zero likelihood. So, I decided to put the whole thing on the shelf, and for a while, the issue died down.

About a year later, Janet Mertz, a graduate student in my lab, discovered that EcoRI made cohesive ends. She found that she could take any two DNAs and join them together. This breakthrough allowed Stan Cohen and Herb Boyer to begin making recombinants using E. coli plasmids, checking for drug resistance, and showing they could create chimeric DNA molecules.

They presented their work at a Gordon Conference, which sparked a similar reaction among a small minority of the audience. Some expressed alarm, saying, "Oh my God, you guys could really make some dangerous things." At that conference, Maxine Singer and Dieter Soll, the co-chairs, were persuaded to draft a letter to the academy, published in Science. The letter highlighted the great promise of being able to create new kinds of DNA molecules while also addressing the possible dangers and risks.

When the president of the academy received the letter, he asked, "What do I do with it?" Maxine suggested, "Call Paul Berg, because he's thought about this problem."


The papers presented at the 1973 Gordon Conference alarmed the proto-molecular biologist conferees, who voted to send a letter to Philip Handler, President of the National Academy of Sciences, about the possible consequences of the new technology. The letter read:

Guidelines for DNA Hybrid Molecules

Those in attendance at the 1973 Gordon Conference on Nucleic Acids voted to send the following letter to Philip Handler, president of the National Academy of Sciences, and to John R. Hogness, president of the National Institute of Medicine. A majority also desired to publicize the letter more widely.

We are writing to you, on behalf of a number of scientists, to communicate a matter of deep concern. Several of the reports presented at this year's Gordon Research Conference on Nucleic Acids (June 11—15, 1973, New Hampton. New Hampshire) indicated that we presently have the technical ability to join together, covalently, DNA molecules from diverse sources. Scientific developments over the past two years make it both reasonable and convenient to generate overlapping sequence homologies at the termini of different DNA molecules. The sequence homologies can then be used to combine the molecules by Watson-Crick hydrogen bonding. Application of existing methods permits subsequent covalent linkage of such molecules. This technique could be used, for example, to combine DNA from animal viruses with bacterial DNA, or DNA's of different viral origin might be joined. In this way new kinds of hybrid plasmids or viruses, with biological activity of unpredictable nature, may eventually be created. These experiments offer exciting and interesting potential both for advancing knowledge of fundamental biological processes and for alleviation of human health problems.

Certain such hybrid molecules may prove hazardous to laboratory workers and to the public. Although no hazard has yet been established, prudence suggests that the potential hazard be seriously considered. A majority of those attending the Conference voted to communicate their concern in this matter to you and to the President of the Institute of Medicine (to whom this letter is also being sent). The conferees suggested that the Academies establish a study committee to consider this problem and to recommend specific actions or guidelines, should that seem appropriate. Related problems such as the risks involved in current large-scale preparation of animal viruses might also be considered.

(signed) MAXINE SINGER  DIETER SOLL

From this initial burst of activity, scientists reached a consensus that the as-yet unnamed science of Genetic Engineering should be put under a moratorium. Experiments were to cease until the ramifications could be discussed and amelioration put into place.

The conference meeting was held at Asilomar, in Southern California, in February 1975. The outcome was a set of guidelines for handling recombinant DNA, physically (biological containment facilities) and experimentally (using only hosts that would die outside a laboratory). 

That was 50 years ago.

I wasn’t a “Genetic Engineer” in those days. I’m old, but not that old. I learned about the conference in 1979, as I studied genetics for my degree. 

The burgeoning field of genetics had a vertical take off between the first “restriction enzymes” in the early seventies and the routine gene engineering of the 21st Century. In fact, the modern understanding of heredity is less than two hundred years old: The understanding that for all sexually reproducing species – plant or animal - two parents provide heritable material to an embryo, which can then pass down half its own heritable material to its offspring.

A brief timeline:

1. Mendel's experiments showing that various plant characteristics were not “blends” of their parents’ characteristics: 1865

2. The concept of evolution set down in writing by Darwin. The Origin of Species, 1859, The Descent of Man published on February 24, 1871

3. Mendel’s experiments verified (and rediscovered) by Hugo de Vries, Carl Erich Correns and von Seysenegg, 1900

4. Proof of DNA's role in heredity:  The Hershey-Chase experiment, which demonstrated that DNA, not protein, was the genetic material, 1952

5. Discovery of DNA structure: The double helix structure of DNA was identified by James Watson and Francis Crick in mid-March 1953, with their paper published on April 25, 1953

6. Experiments showing DNA replication: The Meselson-Stahl experiment providing experimental proof of DNA's semiconservative replication: 1958

7. The first paper describing a restriction enzyme published by Hamilton O. Smith, Thomas Kelly, and Kent Wilcox. The isolation and characterization of the type II restriction enzyme Hind II from Haemophilus influenzae: 1970

8. The Human Genome Project:

Started in 1990

Declared complete on April 14, 2003 (covering about 92% of the genome)

"Complete genome" level achieved in May 2021

Final gapless assembly finished in January 2022

Recombinant technology has moved on at a tremendous rate since the seventies. Several other conferences have been held, including this year at the "Spirit of Asilomar and the Future of Biotechnology" conference in February 2025, and work continues.

The Asilomar conference marked the first significant time that researchers stopped and thought, “Just because I can, it doesn’t mean I should.” Scientific Ethics have been in the news again and again after 1975 and there are very good reasons for that.

Another thing that has been in the news over the past few years, as well: the possibility of “lab leaks.”

What did people in 1974 know about lab leaks?

“This [genetic manipulation] would call for technical re-education of the average microbial geneticist or molecular biologist, whose manipulation of bacteria chills the blood of anyone accustomed to handling pathogens.” (Emphasis mine.)

Thank you, Professor E. S. Anderson of Colindale. [Nature (250(5464), 278–280. doi:10.1038/250278b0]

For background, reading, please see Paul Berg's video and/or my earlier posts on Asilomar,

https://peromyscus.blogspot.com/2013/06/genetics-library-project-political.html

https://peromyscus.blogspot.com/2015/03/genetic-engineering-birds-do-it-bees-do.html

LinkWithin

Blog Widget by LinkWithin
I sometimes mention a product on this blog, and I give a URL to Amazon or similar sites. Just to reassure you, I don't get paid to advertise anything here and I don't get any money from your clicks. Everything I say here is because I feel like saying it.