Pinned Tweet
A weekly jab in the belly is generating more revenue than the entire AI industry.
Ozempic + Mounjaro: $71B in 2025.
OpenAI + Anthropic: $29B.
And they've barely started. ~2% of the 800 million eligible patients can currently access them.
h/t @DrSamuelBHume
Ozempic + Mounjaro: $71B in 2025.
OpenAI + Anthropic: $29B.
And they've barely started. ~2% of the 800 million eligible patients can currently access them.
h/t @DrSamuelBHume
79
253
1,503
869,668
If you are over 60 and taking a daily aspirin to prevent cancer, the biggest trial in older adults delivered a jarring result: more cancer deaths on aspirin than on placebo.
ASPREE followed 19,114 Australians and Americans aged 70 and older, 65 and up for some participants, who took 100 mg of aspirin or placebo for about five years. Cancer deaths were 3.1% with aspirin versus 2.3% with placebo. Major bleeding rose by more than a third. New cancer diagnoses were about the same, and the study’s authors called the death finding unexpected.
Other evidence points the other way. In 861 people with Lynch syndrome, an inherited high-risk condition, 600 mg a day for up to four years cut colorectal cancer from 13% to 9% over ten years. After surgery for colorectal cancer with PIK3CA changes, 160 mg lowered three-year recurrence from 14 in 100 to 8. Peter Rothwell’s pooled analysis of earlier heart trials, mostly in younger people, found about a third fewer cancer deaths, but only after five years.
The bottom line: aspirin is not a simple cancer-prevention play for older adults. The USPSTF advises against starting it at age 60 or older to prevent a first heart attack or stroke.
ASPREE followed 19,114 Australians and Americans aged 70 and older, 65 and up for some participants, who took 100 mg of aspirin or placebo for about five years. Cancer deaths were 3.1% with aspirin versus 2.3% with placebo. Major bleeding rose by more than a third. New cancer diagnoses were about the same, and the study’s authors called the death finding unexpected.
Other evidence points the other way. In 861 people with Lynch syndrome, an inherited high-risk condition, 600 mg a day for up to four years cut colorectal cancer from 13% to 9% over ten years. After surgery for colorectal cancer with PIK3CA changes, 160 mg lowered three-year recurrence from 14 in 100 to 8. Peter Rothwell’s pooled analysis of earlier heart trials, mostly in younger people, found about a third fewer cancer deaths, but only after five years.
The bottom line: aspirin is not a simple cancer-prevention play for older adults. The USPSTF advises against starting it at age 60 or older to prevent a first heart attack or stroke.
Daily aspirin has been linked to lower rates of colorectal, oesophageal, stomach, liver, breast, lung, prostate, ovarian, pancreatic, endometrial, skin and head and neck cancers, melanoma, leukaemia and some brain cancers, as well as fewer heart attacks, certain strokes, recurrent blood clots, colorectal polyps and pre-eclampsia. It’s also been used for migraines, headaches, toothaches, back pain, arthritis, sore joints, muscle pain, menstrual cramps, fever, inflammation and even reducing facial flushing. And then there are people reporting improvements in things like acne, rosacea, skin irritation, period symptoms, sleep, mood, energy, brain fog and feeling generally less sore and inflamed. There’s also research into whether it might help with cancer spreading, dementia and other diseases associated with ageing. All from a little tablet…that’s been around for well over 100 years.
6
8
37
14,880
Why do cancer scan results hit your app at 11 PM, before a doctor calls?
That Friday-night PDF full of "suspicious" and "cannot be excluded” has nearly a million views for a reason.
It is a federal rule. Since 2021, the 21st Century Cures Act has treated routinely holding results until a doctor reviews them as likely "information blocking."
Mayo Clinic used to wait 36 hours. After it stopped, patients saw their scan report before the ordering doctor 44% of the time, up from 18.5%, across 1.2 million scans.
Most patients want instant release. In one survey of 8,139 patients at four US hospitals, 95.7% preferred it. But 16.5% of those with abnormal results said seeing them first made them more worried.
If you want a person first, ask your doctor's office to hold the results until someone calls. Federal guidance says that is not information blocking.
That Friday-night PDF full of "suspicious" and "cannot be excluded” has nearly a million views for a reason.
It is a federal rule. Since 2021, the 21st Century Cures Act has treated routinely holding results until a doctor reviews them as likely "information blocking."
Mayo Clinic used to wait 36 hours. After it stopped, patients saw their scan report before the ordering doctor 44% of the time, up from 18.5%, across 1.2 million scans.
Most patients want instant release. In one survey of 8,139 patients at four US hospitals, 95.7% preferred it. But 16.5% of those with abnormal results said seeing them first made them more worried.
If you want a person first, ask your doctor's office to hold the results until someone calls. Federal guidance says that is not information blocking.
Healthcare is the only industry that drops your cancer scan results into an app at 11 PM on a Friday, then closes for the weekend.
No call. No context. Just a PDF full of words like "suspicious" and "cannot be excluded" and a little banner that says "Sleep well!"
Meanwhile ours 😐
2
1
15
2,216
Can you still smell coffee as well as you used to? Scientists at Ohio State just built a gadget that sends smells over the internet, and their paper says it could also test exactly that.
It's a test worth taking. A University of Chicago team gave about 3,000 Americans aged 57 to 85 a three-minute sniff test with five scented pens.
Five years later, 39% of those who missed four or five scents had died, against 10% of those who missed none or one. The gap held after accounting for age, cancer and heart failure.
Losing smell probably doesn't cause death. The nose rebuilds its smell cells from stem cells, so a fading nose may be an early sign that the body's repair is slowing down.
If scents are fading for you or a parent, ask about a quick smell test at your next checkup.
It's a test worth taking. A University of Chicago team gave about 3,000 Americans aged 57 to 85 a three-minute sniff test with five scented pens.
Five years later, 39% of those who missed four or five scents had died, against 10% of those who missed none or one. The gap held after accounting for age, cancer and heart failure.
Losing smell probably doesn't cause death. The nose rebuilds its smell cells from stem cells, so a fading nose may be an early sign that the body's repair is slowing down.
If scents are fading for you or a parent, ask about a quick smell test at your next checkup.
BREAKING: Scientists develop technology that allows people to digitally transmit smells across the world, potentially bringing scent to phone calls, video chats, & social media.
1
1
8
780
Do cancer vaccines work yet? In the past two weeks, three trials have given three very different answers.
Yes, in melanoma. Moderna and Merck built an mRNA vaccine from each patient's tumor mutations after surgery, then gave it with Keytruda. In 1,137 patients, 82% were alive ce at two years, versus 70% on Keytruda alone. The full data were presented today at ESMO in Madrid.
Not in pancreatic cancer, at least not yet. Elicio's off-the-shelf vaccine targets seven common mutations in the KRAS cancer gene and was given after surgery and standard treatment. In 144 patients, it did not clearly reduce recurrence overall, though some subgroups did better, and the company is talking to the FDA about a phase 3.
Not yet, either, in advanced head and neck cancer. BioNTech's vaccine against two proteins from the HPV16 virus, added to Keytruda, did not clearly slow the cancer, and the trial now waits to see whether people live longer.
The pattern matters. The vaccine that worked was personalized to each patient’s tumor and used after surgery had removed the cancer doctors could see. Even then, it took about six months before the two groups pulled apart.
The clearest signal so far is this: a patient-specific vaccine can help keep melanoma from coming back after surgery. The next question is which cancers can follow.
Yes, in melanoma. Moderna and Merck built an mRNA vaccine from each patient's tumor mutations after surgery, then gave it with Keytruda. In 1,137 patients, 82% were alive ce at two years, versus 70% on Keytruda alone. The full data were presented today at ESMO in Madrid.
Not in pancreatic cancer, at least not yet. Elicio's off-the-shelf vaccine targets seven common mutations in the KRAS cancer gene and was given after surgery and standard treatment. In 144 patients, it did not clearly reduce recurrence overall, though some subgroups did better, and the company is talking to the FDA about a phase 3.
Not yet, either, in advanced head and neck cancer. BioNTech's vaccine against two proteins from the HPV16 virus, added to Keytruda, did not clearly slow the cancer, and the trial now waits to see whether people live longer.
The pattern matters. The vaccine that worked was personalized to each patient’s tumor and used after surgery had removed the cancer doctors could see. Even then, it took about six months before the two groups pulled apart.
The clearest signal so far is this: a patient-specific vaccine can help keep melanoma from coming back after surgery. The next question is which cancers can follow.
8
4
61
6,069
Sources and Notes
1. MELANOMA, THE WIN
- INTerpath-001, Moderna and Merck, presented October 24, 2026 at the ESMO Congress in Madrid (Presidential Symposium). 1,137 people with stage IIB to IV melanoma removed by surgery got intismeran autogene, an mRNA vaccine built from each person's tumor mutations, plus Keytruda (pembrolizumab), or placebo plus Keytruda, for about a year.
merck.com/news/merck-and…
- At two years, 82.2% were alive without the cancer coming back on the combination, against 70.0% on Keytruda alone (hazard ratio 0.58). Spread to distant organs was also less common (hazard ratio 0.62). Overall survival isn't mature yet.
- For the first six months the groups looked the same (hazard ratio 0.99); the benefit appeared after that. Severe (grade 3 or higher) treatment-related side effects were similar, 14.2% against 13.4%.
· · · · ·
2. PANCREATIC CANCER, NOT OVERALL
- AMPLIFY-7P, Elicio Therapeutics, a phase 2 trial presented at ESMO 2026 (published in Nature Medicine). 144 people with KRAS-mutant pancreatic cancer, after surgery and standard treatment, got an off-the-shelf vaccine against seven common KRAS mutations or observation.
globenewswire.com/news-release/2…
- Disease-free survival wasn't clearly better overall (hazard ratio 0.85). Subgroups, including people whose blood showed no tumor DNA and people who hadn't had radiation, did better, but those splits need a new trial to confirm. The company plans to discuss the findings with the FDA to inform a possible phase 3.
· · · · ·
3. HEAD AND NECK CANCER, NOT YET
- AHEAD-MERIT, BioNTech, update of October 13, 2026. A phase 2/3 trial in advanced, HPV16-positive head and neck cancer that had spread or come back: BNT113, an mRNA vaccine against the HPV16 proteins E6 and E7, plus pembrolizumab, against pembrolizumab alone.
globenewswire.com/news-release/2…
- It missed its first main goal, slowing the cancer (progression-free survival). The trial continues to its second main goal, overall survival.
· · · · ·
4. MY READ
- Cancer vaccines aren't new: sipuleucel-T (Provenge) was approved for advanced prostate cancer in 2010. INTerpath-001 is the first phase 3 win for a vaccine built from each patient's own tumor.
- The three trials differ in more than one way: cancer type, how advanced it was, whether the vaccine was personalized, and trial size. So the pattern in the post (personalized, after surgery) is my reading of three trials and still needs testing.
1. MELANOMA, THE WIN
- INTerpath-001, Moderna and Merck, presented October 24, 2026 at the ESMO Congress in Madrid (Presidential Symposium). 1,137 people with stage IIB to IV melanoma removed by surgery got intismeran autogene, an mRNA vaccine built from each person's tumor mutations, plus Keytruda (pembrolizumab), or placebo plus Keytruda, for about a year.
merck.com/news/merck-and…
- At two years, 82.2% were alive without the cancer coming back on the combination, against 70.0% on Keytruda alone (hazard ratio 0.58). Spread to distant organs was also less common (hazard ratio 0.62). Overall survival isn't mature yet.
- For the first six months the groups looked the same (hazard ratio 0.99); the benefit appeared after that. Severe (grade 3 or higher) treatment-related side effects were similar, 14.2% against 13.4%.
· · · · ·
2. PANCREATIC CANCER, NOT OVERALL
- AMPLIFY-7P, Elicio Therapeutics, a phase 2 trial presented at ESMO 2026 (published in Nature Medicine). 144 people with KRAS-mutant pancreatic cancer, after surgery and standard treatment, got an off-the-shelf vaccine against seven common KRAS mutations or observation.
globenewswire.com/news-release/2…
- Disease-free survival wasn't clearly better overall (hazard ratio 0.85). Subgroups, including people whose blood showed no tumor DNA and people who hadn't had radiation, did better, but those splits need a new trial to confirm. The company plans to discuss the findings with the FDA to inform a possible phase 3.
· · · · ·
3. HEAD AND NECK CANCER, NOT YET
- AHEAD-MERIT, BioNTech, update of October 13, 2026. A phase 2/3 trial in advanced, HPV16-positive head and neck cancer that had spread or come back: BNT113, an mRNA vaccine against the HPV16 proteins E6 and E7, plus pembrolizumab, against pembrolizumab alone.
globenewswire.com/news-release/2…
- It missed its first main goal, slowing the cancer (progression-free survival). The trial continues to its second main goal, overall survival.
· · · · ·
4. MY READ
- Cancer vaccines aren't new: sipuleucel-T (Provenge) was approved for advanced prostate cancer in 2010. INTerpath-001 is the first phase 3 win for a vaccine built from each patient's own tumor.
- The three trials differ in more than one way: cancer type, how advanced it was, whether the vaccine was personalized, and trial size. So the pattern in the post (personalized, after surgery) is my reading of three trials and still needs testing.
0
0
2
614
A cattle virus that appeared in Europe this summer has now killed its first known human.
The patient was in their sixties, from rural eastern France, and had leukemia. After treatment and a stem cell transplant, they no longer had B cells to make antibodies. On Sept. 23, they went blind and confused, then fell into a coma.
Routine tests found nothing. Doctors in Lyon sequenced the spinal fluid and identified Shamonda, a virus never before seen in a person. The patient died on Oct. 9.
Stored blood showed the virus had been present since at least July 2, nearly three months before symptoms began.
By Oct. 15, cattle outbreaks had been confirmed in 53 of mainland France's 96 departments. The virus is thought to spread by biting midges.
The key point: this is one case in a severely immunocompromised patient. The doctors say it is not evidence of a serious risk to healthy people, but they want Europe to start tracking the virus in humans as well as animals.
The patient was in their sixties, from rural eastern France, and had leukemia. After treatment and a stem cell transplant, they no longer had B cells to make antibodies. On Sept. 23, they went blind and confused, then fell into a coma.
Routine tests found nothing. Doctors in Lyon sequenced the spinal fluid and identified Shamonda, a virus never before seen in a person. The patient died on Oct. 9.
Stored blood showed the virus had been present since at least July 2, nearly three months before symptoms began.
By Oct. 15, cattle outbreaks had been confirmed in 53 of mainland France's 96 departments. The virus is thought to spread by biting midges.
The key point: this is one case in a severely immunocompromised patient. The doctors say it is not evidence of a serious risk to healthy people, but they want Europe to start tracking the virus in humans as well as animals.
A cancer patient in France has become the first human known to contract the Shamonda virus
go.nature.com/4dWWzGe
3
2
21
1,531
Cells from people as old as 101 were turned into young-looking brain stem cells on DNA aging clocks. And it happened in a dish, not in a person.
A University of Bonn team led by Oliver Brüstle took blood from 15 donors, newborn to 101, and used two genes to turn the cells directly into neural stem cells, the cells that become neurons. In about 50 days, the cells' epigenetic age, measured from chemical marks on their DNA, fell to between a twentieth and a fifth of the donor's age.
The reset itself is not new. Turning cells into stem cells has pushed these clocks close to zero since at least 2013. What is new is that this route is slow, which let the team watch the clock run backward step by step.
This does not make a person younger. It gives scientists a way to see what the reset changes inside a cell, and to ask whether a gentler version could ever work in the body.
A University of Bonn team led by Oliver Brüstle took blood from 15 donors, newborn to 101, and used two genes to turn the cells directly into neural stem cells, the cells that become neurons. In about 50 days, the cells' epigenetic age, measured from chemical marks on their DNA, fell to between a twentieth and a fifth of the donor's age.
The reset itself is not new. Turning cells into stem cells has pushed these clocks close to zero since at least 2013. What is new is that this route is slow, which let the team watch the clock run backward step by step.
This does not make a person younger. It gives scientists a way to see what the reset changes inside a cell, and to ask whether a gentler version could ever work in the body.
Next step: Longevity and Rejuvenation. We are almost there.
Researchers in Bonn turned blood cells from an 80-year-old into neural stem cells with a molecular age below 20. More than 60 years younger by epigenetic clocks, which estimate age from chemical marks on DNA.
The team lood precursor cells into neural stem cells, skipping the usual pluripotent stem cell stage. The rejuvenation unfolded gradually and could be tracked for over 100 days.
1
2
15
1,182
Millions of people take Ozempic or Mounjaro, and many wonder if these drugs might also protect memory. So far, the biggest test says no. In 3,808 people with early Alzheimer's, the drug in Ozempic did not slow the disease.
The idea came from health records. A 2024 Case Western with type 2 diabetes on semaglutide, the drug in Ozempic and Wegovy, were diagnosed with Alzheimer's about a third as often as similar people on insulin.
Novo Nordisk, the maker of Ozempic, then ran a proper trial. In two studies, people took semaglutide or a placebo for two years. Their memory and daily life declined at the same pace.
These results answer different questions. Health records can be skewed because people on insulin are often sicker to begin with, while the trials tested people who already had Alzheimer's. Mounjaro has no Alzheimer's trial results yet.
The bottom line: these drugs still help with weight, blood sugar, and in the case of Ozempic, heart attacks and strokes. But for memory, the largest trials so far came up empty.
The idea came from health records. A 2024 Case Western with type 2 diabetes on semaglutide, the drug in Ozempic and Wegovy, were diagnosed with Alzheimer's about a third as often as similar people on insulin.
Novo Nordisk, the maker of Ozempic, then ran a proper trial. In two studies, people took semaglutide or a placebo for two years. Their memory and daily life declined at the same pace.
These results answer different questions. Health records can be skewed because people on insulin are often sicker to begin with, while the trials tested people who already had Alzheimer's. Mounjaro has no Alzheimer's trial results yet.
The bottom line: these drugs still help with weight, blood sugar, and in the case of Ozempic, heart attacks and strokes. But for memory, the largest trials so far came up empty.
6
1
15
1,038
Sources and Notes
1. THE HEALTH-RECORDS STUDY
- Xu R and colleagues, Case Western Reserve University. Seven emulated target trials in 1,094,761 US adults with type 2 diabetes and no prior Alzheimer's diagnosis, followed for 3 years. Semaglutide versus insulin: hazard ratio 0.33 (95% CI 0.21 to 0.51) for a first Alzheimer's diagnosis. Versus other GLP-1 drugs: 0.59 (0.37 to 0.95). Observational, so it shows a link, not cause.
doi.org/10.1002/alz.14…
2. THE TRIALS (evoke and evoke+)
- 3,808 people aged 55 to 85 with amyloid-confirmed mild cognitive impairment or mild Alzheimer's dementia, randomized to oral semaglutide 14 mg or placebo for up to 156 weeks at 566 sites in 40 countries. Most did not have diabetes.
- Change in CDR-SB at week 104, semaglutide versus placebo: evoke 2.3 versus 2.3 (difference -0.08, 95% CI -0.35 to 0.20); evoke+ 2.2 versus 2.1 (difference +0.10, -0.17 to 0.38). The authors conclude it was not efficacious.
doi.org/10.1016/S0140-…
3. THE POOLED ANALYSIS
- Meta-analysis of seven randomized trials, 4,132 participants with Alzheimer's or mild cognitive impairment, about 92% of them from evoke and evoke+: no cognitive benefit and no slowing on CDR-SB.
doi.org/10.1002/trc2.7…
- The liraglutide ELAD trial (204 people, 52 weeks) also missed its primary outcome.
doi.org/10.1038/s41591…
4. WHY THE TWO KINDS OF EVIDENCE CAN DISAGREE
- The records study asked whether people with diabetes on the drug were diagnosed less often. The trials asked whether the drug slows Alzheimer's once it has begun, mostly in people without diabetes. The records study matched patients on measured characteristics, but unmeasured differences between people prescribed insulin and people prescribed semaglutide can remain. No randomized trial has tested whether semaglutide prevents Alzheimer's.
1. THE HEALTH-RECORDS STUDY
- Xu R and colleagues, Case Western Reserve University. Seven emulated target trials in 1,094,761 US adults with type 2 diabetes and no prior Alzheimer's diagnosis, followed for 3 years. Semaglutide versus insulin: hazard ratio 0.33 (95% CI 0.21 to 0.51) for a first Alzheimer's diagnosis. Versus other GLP-1 drugs: 0.59 (0.37 to 0.95). Observational, so it shows a link, not cause.
doi.org/10.1002/alz.14…
2. THE TRIALS (evoke and evoke+)
- 3,808 people aged 55 to 85 with amyloid-confirmed mild cognitive impairment or mild Alzheimer's dementia, randomized to oral semaglutide 14 mg or placebo for up to 156 weeks at 566 sites in 40 countries. Most did not have diabetes.
- Change in CDR-SB at week 104, semaglutide versus placebo: evoke 2.3 versus 2.3 (difference -0.08, 95% CI -0.35 to 0.20); evoke+ 2.2 versus 2.1 (difference +0.10, -0.17 to 0.38). The authors conclude it was not efficacious.
doi.org/10.1016/S0140-…
3. THE POOLED ANALYSIS
- Meta-analysis of seven randomized trials, 4,132 participants with Alzheimer's or mild cognitive impairment, about 92% of them from evoke and evoke+: no cognitive benefit and no slowing on CDR-SB.
doi.org/10.1002/trc2.7…
- The liraglutide ELAD trial (204 people, 52 weeks) also missed its primary outcome.
doi.org/10.1038/s41591…
4. WHY THE TWO KINDS OF EVIDENCE CAN DISAGREE
- The records study asked whether people with diabetes on the drug were diagnosed less often. The trials asked whether the drug slows Alzheimer's once it has begun, mostly in people without diabetes. The records study matched patients on measured characteristics, but unmeasured differences between people prescribed insulin and people prescribed semaglutide can remain. No randomized trial has tested whether semaglutide prevents Alzheimer's.
0
0
1
221
A personalized cancer vaccine just delivered strong results in a large trial, yet Moderna's stock still fell more than 17% the same day.
Scientists sequence each patient's melanoma after surgery, create an mRNA vaccine from the mutations, and combine it with Keytruda. In a phase 3 trial of 1,137 patients, the risk of recurrence or death dropped 42%. After two years, 82% of patients were alive with no recurrence, compared with 70% in the Keytruda plus placebo group.
The catch is that the data leaked two days before the presentation, and some oncologists say the bar is higher than it appears. In a 2023 trial, Keytruda alone before and after surgery cut the risk of recurrence or worsening by a similar amount. The patients and setup were different, but the vaccine now has to show it adds more than a well-established drug on its own.
It is not approved yet, it is not clear whether it helps people live longer, and the full data will be available tomorrow (Saturday) at the ESMO cancer meeting in Madrid.
Scientists sequence each patient's melanoma after surgery, create an mRNA vaccine from the mutations, and combine it with Keytruda. In a phase 3 trial of 1,137 patients, the risk of recurrence or death dropped 42%. After two years, 82% of patients were alive with no recurrence, compared with 70% in the Keytruda plus placebo group.
The catch is that the data leaked two days before the presentation, and some oncologists say the bar is higher than it appears. In a 2023 trial, Keytruda alone before and after surgery cut the risk of recurrence or worsening by a similar amount. The patients and setup were different, but the vaccine now has to show it adds more than a well-established drug on its own.
It is not approved yet, it is not clear whether it helps people live longer, and the full data will be available tomorrow (Saturday) at the ESMO cancer meeting in Madrid.
BREAKING: $MRNA has crashed 25% today as its cancer vaccine data leaked online before Saturday's presentation.
The leaked data reportedly shows a 42% lower risk of recurrence or death vs Keytruda alone, but the stock is still crashing.
Just yesterday, the stock hit its highest .
6
3
40
4,091
At 35, you may feel young, but the arteries feeding your heart may already be changing almost twice as fast as they will at 55.
That comes from a study of more than 25,000 microscope slides from 970 adults ages 21 to 70. Researchers at Sanford Burnham Prebys and the NIH used an 40 tissues change with age.
The coronary arteries changed fastest in the 30s. Pathologists' notes also showed early plaque rising most sharply between the 20s and 30s.
The bigger lesson: organs do not age on the same schedule. The uterus and vagina changed most in the 50s, around menopause.
This was an average across people, since each donor was seen only once, after death. Brain arteries were not included.
Your body does not age all at once. Your arteries start earlier than you think
That comes from a study of more than 25,000 microscope slides from 970 adults ages 21 to 70. Researchers at Sanford Burnham Prebys and the NIH used an 40 tissues change with age.
The coronary arteries changed fastest in the 30s. Pathologists' notes also showed early plaque rising most sharply between the 20s and 30s.
The bigger lesson: organs do not age on the same schedule. The uterus and vagina changed most in the 50s, around menopause.
This was an average across people, since each donor was seen only once, after death. Brain arteries were not included.
Your body does not age all at once. Your arteries start earlier than you think
9
4
37
1,788
Sources and Notes
1. THE PAPER
- Yadav, Sinha and colleagues at Sanford Burnham Prebys, the National Cancer Institute and the NICHD, Nature Aging, published 31 August 2026. They built PathStAR, which scores how tissue structure changes with age from routine pathology slides, without being trained to predict age.
doi.org/10.1038/s43587…
- Free full text. The image uses Fig. 4d (coronary artery against tibial nerve), CC BY 4.0, figure credited in part to BioRender (Alvarez, K.).
pmc.ncbi.nlm.nih.gov/articles/PMC13…
· · · · ·
2. THE DONORS
- 25,306 post-mortem biopsy slides from 40 tissues in 970 donors aged 21 to 70, from the US GTEx tissue bank. 16,808 slides came from men and 8,498 from women, so about a third from women. Most samples come from donors over 45.
pubmed.ncbi.nlm.nih.gov/42675190/
· · · · ·
3. WHAT THEY FOUND
- Vascular tissue, including the tibial and coronary arteries, had its highest structural aging rate in the 30s. In Fig. 4d the coronary rate at age 35 is about 0.24, almost twice its rate at 55 (about 0.13). The curves cover ages 30 to 60 and 15 of the 40 tissues.
- The pattern held after removing donors with systemic diseases. In the pathology notes, early plaque rose most steeply from the 20s to the 30s, then plateaued.
- The uterus and vagina changed fastest in the early and middle 50s, around menopause. The ovary had two peaks, at 35 to 40 and at 55 to 60.
- A 2025 Cell study of 516 samples from 13 tissues, which measured proteins and also examined tissue histology, found blood vessels among the tissues that age early.
doi.org/10.1016/j.cell…
· · · · ·
4. LIMITS
- Each donor was sampled once, after death, so these are averages across people rather than one person followed over time.
- The score measures change in structure. The authors say some change may be adaptation or neutral, and the paper doesn't test how well the arteries work.
- Ages are pooled in 10-year windows, and GTEx donors vary in health, ventilator use and time to sampling. Post-mortem factors explained under 10% of the rates in almost all tissues.
- Brain arteries weren't studied. Two authors, Yadav and Sinha, are named inventors on a related US provisional patent application. Funding came from a Sanford Burnham Prebys start-up fund and the NIH Intramural Research Program.
1. THE PAPER
- Yadav, Sinha and colleagues at Sanford Burnham Prebys, the National Cancer Institute and the NICHD, Nature Aging, published 31 August 2026. They built PathStAR, which scores how tissue structure changes with age from routine pathology slides, without being trained to predict age.
doi.org/10.1038/s43587…
- Free full text. The image uses Fig. 4d (coronary artery against tibial nerve), CC BY 4.0, figure credited in part to BioRender (Alvarez, K.).
pmc.ncbi.nlm.nih.gov/articles/PMC13…
· · · · ·
2. THE DONORS
- 25,306 post-mortem biopsy slides from 40 tissues in 970 donors aged 21 to 70, from the US GTEx tissue bank. 16,808 slides came from men and 8,498 from women, so about a third from women. Most samples come from donors over 45.
pubmed.ncbi.nlm.nih.gov/42675190/
· · · · ·
3. WHAT THEY FOUND
- Vascular tissue, including the tibial and coronary arteries, had its highest structural aging rate in the 30s. In Fig. 4d the coronary rate at age 35 is about 0.24, almost twice its rate at 55 (about 0.13). The curves cover ages 30 to 60 and 15 of the 40 tissues.
- The pattern held after removing donors with systemic diseases. In the pathology notes, early plaque rose most steeply from the 20s to the 30s, then plateaued.
- The uterus and vagina changed fastest in the early and middle 50s, around menopause. The ovary had two peaks, at 35 to 40 and at 55 to 60.
- A 2025 Cell study of 516 samples from 13 tissues, which measured proteins and also examined tissue histology, found blood vessels among the tissues that age early.
doi.org/10.1016/j.cell…
· · · · ·
4. LIMITS
- Each donor was sampled once, after death, so these are averages across people rather than one person followed over time.
- The score measures change in structure. The authors say some change may be adaptation or neutral, and the paper doesn't test how well the arteries work.
- Ages are pooled in 10-year windows, and GTEx donors vary in health, ventilator use and time to sampling. Post-mortem factors explained under 10% of the rates in almost all tissues.
- Brain arteries weren't studied. Two authors, Yadav and Sinha, are named inventors on a related US provisional patent application. Funding came from a Sanford Burnham Prebys start-up fund and the NIH Intramural Research Program.
0
0
4
262
Americans are flying to Shanghai for a lung cancer pill that the FDA approved 15 months ago.
Max Doppelt did it for his mother. Her lung cancer had spread to her brain. The drug most likely to help her, sunvozertinib, had been sold in China since 2023 but was unavailable in the came home with 26 boxes. Nineteen months later, she is alive and still working. He says at least 30 Americans have made the same trip.
In the trial behind the US approval, the pill shrank tumors in 46 out of every 100 patients whose cancer had worsened after chemotherapy.
The holdup is business, not medicine. Dizal, the Shanghai company that makes it, had no US sales network, according to US oncologists in May. AstraZeneca licensed the drug in July for $600 million and says US sales will start this quarter.
An @FDA approval helps American patients only when a company actually sells the drug here.
Max Doppelt did it for his mother. Her lung cancer had spread to her brain. The drug most likely to help her, sunvozertinib, had been sold in China since 2023 but was unavailable in the came home with 26 boxes. Nineteen months later, she is alive and still working. He says at least 30 Americans have made the same trip.
In the trial behind the US approval, the pill shrank tumors in 46 out of every 100 patients whose cancer had worsened after chemotherapy.
The holdup is business, not medicine. Dizal, the Shanghai company that makes it, had no US sales network, according to US oncologists in May. AstraZeneca licensed the drug in July for $600 million and says US sales will start this quarter.
An @FDA approval helps American patients only when a company actually sells the drug here.
In January 2025 my mom’s lung cancer spread to her brain. The drug with the best chance of helping her was approved in China but not here.
A doctor told me, almost as an aside, that if she were Chinese she could buy it at a pharmacy.
So I flew to Shanghai and came home with 26 boxes in my luggage. That was 19 months ago. My mom is still here, still working, and in her words has had an amazing three years.
The drug was approved in China in 2023. The FDA approved it in July 2025. It still is not for sale here. I know of at least 30 Americans who have made the same trip.
I have been scared to tell this story because I worried it could cost my mom access to the medicine keeping her alive. But the problem matters more than my silence. These drugs need to be developed and available here, and that means FDA reform.
Art Allen at KFF Health News told our story this week. It also ran in The Washington Post.
kffhealthnews.org/health-industr…
19
109
505
37,778
Sources and Notes
1. THE DRUG
- KFF Health News (20 Oct 2026) names the drug Max brought back from Shanghai: sunvozertinib (Zegfrovy), made by Dizal. Max told KFF he knows of at least 30 Americans who have made the trip.
kffhealthnews.org/health-industr…
- China approved it on 23 Aug 2023 for lung cancer with EGFR exon 20 insertion mutations after platinum chemotherapy.
biospace.com/dizal-s-sunvoz…
- FDA accelerated approval, 2 Jul 2025: confirmed response rate 46% (95% CI 35 to 57) in 85 patients on 200 mg, median response 11.1 months (WU-KONG1B).
fda.gov/drugs/resource…
2. WHY IT ISN'T SOLD HERE YET
- Ou, Nagasaka and colleagues (May 2026) describe it as approved but largely unavailable in the US, citing no US commercial presence, China-based manufacturing and tariffs.
dovepress.com/sunvozertinib-…
- AstraZeneca licensed worldwide rights in July 2026 ($600M upfront, up to $900M in milestones; closed 1 Sep) and says it will launch in the US in Q4 2026. An AstraZeneca spokesperson told KFF availability is expected by the start of 2027.
astrazeneca.com/media-centre/p…
3. BRINGING DRUGS HOME
- FDA's personal importation policy allows discretion for up to a 3-month supply of a drug for a serious condition when no effective treatment is available in the US, with a US doctor named. It isn't a license.
fda.gov/industry/impor…
4. LIMITS
- The approval is accelerated, based on tumor response in a single-arm trial. A phase 3 first-line trial (WU-KONG28, NEJM) showed progression-free survival of 10.3 vs 7.5 months against chemotherapy.
doi.org/10.1056/NEJMoa…
1. THE DRUG
- KFF Health News (20 Oct 2026) names the drug Max brought back from Shanghai: sunvozertinib (Zegfrovy), made by Dizal. Max told KFF he knows of at least 30 Americans who have made the trip.
kffhealthnews.org/health-industr…
- China approved it on 23 Aug 2023 for lung cancer with EGFR exon 20 insertion mutations after platinum chemotherapy.
biospace.com/dizal-s-sunvoz…
- FDA accelerated approval, 2 Jul 2025: confirmed response rate 46% (95% CI 35 to 57) in 85 patients on 200 mg, median response 11.1 months (WU-KONG1B).
fda.gov/drugs/resource…
2. WHY IT ISN'T SOLD HERE YET
- Ou, Nagasaka and colleagues (May 2026) describe it as approved but largely unavailable in the US, citing no US commercial presence, China-based manufacturing and tariffs.
dovepress.com/sunvozertinib-…
- AstraZeneca licensed worldwide rights in July 2026 ($600M upfront, up to $900M in milestones; closed 1 Sep) and says it will launch in the US in Q4 2026. An AstraZeneca spokesperson told KFF availability is expected by the start of 2027.
astrazeneca.com/media-centre/p…
3. BRINGING DRUGS HOME
- FDA's personal importation policy allows discretion for up to a 3-month supply of a drug for a serious condition when no effective treatment is available in the US, with a US doctor named. It isn't a license.
fda.gov/industry/impor…
4. LIMITS
- The approval is accelerated, based on tumor response in a single-arm trial. A phase 3 first-line trial (WU-KONG28, NEJM) showed progression-free survival of 10.3 vs 7.5 months against chemotherapy.
doi.org/10.1056/NEJMoa…
0
2
7
1,423
Most of the cholesterol in your blood test never reaches your brain. The brain makes its own and holds about a quarter of all the cholesterol in your body.
That matters because this post is not really about a blood number. It is about where brain cholesterol goes and what happens when that system breaks.
In a Scripps-led study, researchers switched off single genes in millions of mouse brain cells, and an OpenAI model helped interpret the results. When DCTN1 was switched off, neurons increased the genes they use to make cholesterol.
DCTN1 is the gene disrupted in Perry syndrome, a rare inherited brain disease with parkinsonism, and the model pointed to a simple explanation: cholesterol delivery inside those neurons may have failed, so the cells tried to make more.
That fits older brain biology. Astrocytes make cholesterol for neurons and deliver it through APOE. In people of European ancestry, one copy of APOE4 roughly triples the risk of Alzheimer’s disease, while two copies raise it about 15-fold.
The DCTN1 idea is still only a preprint, with no lab test yet. But if it holds up, the takeaway is clear: in the brain, cholesterol is not just about how much there is. It is about where it goes.
That matters because this post is not really about a blood number. It is about where brain cholesterol goes and what happens when that system breaks.
In a Scripps-led study, researchers switched off single genes in millions of mouse brain cells, and an OpenAI model helped interpret the results. When DCTN1 was switched off, neurons increased the genes they use to make cholesterol.
DCTN1 is the gene disrupted in Perry syndrome, a rare inherited brain disease with parkinsonism, and the model pointed to a simple explanation: cholesterol delivery inside those neurons may have failed, so the cells tried to make more.
That fits older brain biology. Astrocytes make cholesterol for neurons and deliver it through APOE. In people of European ancestry, one copy of APOE4 roughly triples the risk of Alzheimer’s disease, while two copies raise it about 15-fold.
The DCTN1 idea is still only a preprint, with no lab test yet. But if it holds up, the takeaway is clear: in the brain, cholesterol is not just about how much there is. It is about where it goes.
GPT-Rosalind surfaced a new lead for drug discovery: could restoring cholesterol transport protect neurons from degeneration?
DCTN1 is a gene linked to Perry syndrome, a rare neurodegenerative disorder. In mouse-brain data, disrupting it increased expression of genes involved in making cholesterol. GPT-Rosalind suggested a possible explanation: cholesterol may not be reaching where neurons need it.
That gives scientists a specific mechanism to test and, if confirmed, a direction for developing new treatments for neurodegenerative disorders.
With @perturbai_tx and @scrippsresearch, a team of thousands of GPT-Rosalind agents worked in parallel, connecting gene-expression results from millions of mouse brain cells with scientific literature in just 32 minutes. We’re now expanding this collaboration to identify new drug targets.
openai.com/index/scaling-…
0
1
22
1,855
Sources and Notes
1. THE TWO PAPERS
- Bilal, Mehra, Rogers and colleagues, Cleveland Clinic and IBM Research, Nature Communications, August 2026. An AI read 9,608 full overnight sleep studies from 9,297 sleep clinic patients and sorted them into five groups, then followed them in medical records. The image is their Figure 2A, CC BY 4.0, with plain labels and notes added.
doi.org/10.1038/s41467…
- Thapa, Mignot, Zou and colleagues, Stanford, Nature Medicine. SleepFM learned from about 585,000 hours of sleep recordings and was tested on 5,019 Stanford sleep clinic patients.
doi.org/10.1038/s41591…
- Rachel Rowe's Nature News and Views on both, 12 October 2026.
doi.org/10.1038/d41586…
· · · · ·
2. WHAT THE AI READ
- Cleveland used 14 channels from in-lab sleep studies, including brain waves (EEG), heart rhythm (ECG), oxygen, airflow and breathing effort. Blanking the brain-wave or heart channels moved many patients into different groups.
· · · · ·
3. THE CLEVELAND NUMBERS
- 7,043 patients and 544 deaths in the death analysis. After adjusting for age, sex, BMI, other illnesses and the breathing-pause count (apnea-hypopnea index), the highest-risk group died at 2.38 times the rate of the lowest (95% CI 1.73 to 3.28). Leaving out patients prescribed CPAP, it was 1.76 (1.12 to 2.78).
- Compared with people with normal counts, people with mild, moderate or severe apnea by that count didn't die more often. Mild (0.71) and moderate (0.52) groups had lower death rates, and severe was 0.77 (0.58 to 1.02).
- The 6 in 100 and 18 in 100 by year 6 are read from Figure 2A, whose curves are weighted so the groups match on age, sex and other factors. Median follow-up was 4.4 years, so the later part of the curves rests on fewer people.
- In the Sleep Heart Health Study, a community group recorded at home, the highest-risk group died at 1.58 times the rate (1.07 to 2.32).
· · · · ·
4. THE STANFORD NUMBERS
- The C-index asks, for two patients where one died sooner, how often the model ranks that one higher. 0.5 is a coin toss. SleepFM, which saw the night plus age and sex, scored 0.84 for death (0.81 to 0.87). A small neural network given only age, sex, BMI and race or ethnicity scored 0.79 (0.75 to 0.82).
- For stroke, the four-fact model scored about the same as SleepFM's 0.78 (Supplementary Figure 11).
· · · · ·
5. LIMITS
- Both main groups were people sent to sleep clinics, who are sicker than the general public. Neither paper tested a watch, ring or headband, and no trial has tested whether acting on these scores helps.
1. THE TWO PAPERS
- Bilal, Mehra, Rogers and colleagues, Cleveland Clinic and IBM Research, Nature Communications, August 2026. An AI read 9,608 full overnight sleep studies from 9,297 sleep clinic patients and sorted them into five groups, then followed them in medical records. The image is their Figure 2A, CC BY 4.0, with plain labels and notes added.
doi.org/10.1038/s41467…
- Thapa, Mignot, Zou and colleagues, Stanford, Nature Medicine. SleepFM learned from about 585,000 hours of sleep recordings and was tested on 5,019 Stanford sleep clinic patients.
doi.org/10.1038/s41591…
- Rachel Rowe's Nature News and Views on both, 12 October 2026.
doi.org/10.1038/d41586…
· · · · ·
2. WHAT THE AI READ
- Cleveland used 14 channels from in-lab sleep studies, including brain waves (EEG), heart rhythm (ECG), oxygen, airflow and breathing effort. Blanking the brain-wave or heart channels moved many patients into different groups.
· · · · ·
3. THE CLEVELAND NUMBERS
- 7,043 patients and 544 deaths in the death analysis. After adjusting for age, sex, BMI, other illnesses and the breathing-pause count (apnea-hypopnea index), the highest-risk group died at 2.38 times the rate of the lowest (95% CI 1.73 to 3.28). Leaving out patients prescribed CPAP, it was 1.76 (1.12 to 2.78).
- Compared with people with normal counts, people with mild, moderate or severe apnea by that count didn't die more often. Mild (0.71) and moderate (0.52) groups had lower death rates, and severe was 0.77 (0.58 to 1.02).
- The 6 in 100 and 18 in 100 by year 6 are read from Figure 2A, whose curves are weighted so the groups match on age, sex and other factors. Median follow-up was 4.4 years, so the later part of the curves rests on fewer people.
- In the Sleep Heart Health Study, a community group recorded at home, the highest-risk group died at 1.58 times the rate (1.07 to 2.32).
· · · · ·
4. THE STANFORD NUMBERS
- The C-index asks, for two patients where one died sooner, how often the model ranks that one higher. 0.5 is a coin toss. SleepFM, which saw the night plus age and sex, scored 0.84 for death (0.81 to 0.87). A small neural network given only age, sex, BMI and race or ethnicity scored 0.79 (0.75 to 0.82).
- For stroke, the four-fact model scored about the same as SleepFM's 0.78 (Supplementary Figure 11).
· · · · ·
5. LIMITS
- Both main groups were people sent to sleep clinics, who are sicker than the general public. Neither paper tested a watch, ring or headband, and no trial has tested whether acting on these scores helps.
0
0
1
358
One night of sleep can hint at how long you might live.
When doctors suspect sleep apnea, they send you for a sleep test. You spend the night wired up in a clinic, and the main score is simple: how many times an hour you stop breathing.
At Cleveland Clinic, that number did not kely to die. An AI that read the rest of the night did. Using brain waves, heartbeat and breathing, it sorted about 9,300 patients into five risk groups.
Six years later, about 18 in 100 in the highest-risk group had died, versus 6 in 100 in the lowest. Even among people the same age and weight, with the same number of breathing stops, the highest-risk group still died at more than twice the rate.
Age, sex and weight do much of the work. At Stanford, a similar AI looked at pairs of patients where one died first and picked the right one 84 times out of 100. A simple model using only age, sex, weight and race got 79.
Both studies looked at people referred to sleep clinics, and neither tested a smartwatch or ring.
When doctors suspect sleep apnea, they send you for a sleep test. You spend the night wired up in a clinic, and the main score is simple: how many times an hour you stop breathing.
At Cleveland Clinic, that number did not kely to die. An AI that read the rest of the night did. Using brain waves, heartbeat and breathing, it sorted about 9,300 patients into five risk groups.
Six years later, about 18 in 100 in the highest-risk group had died, versus 6 in 100 in the lowest. Even among people the same age and weight, with the same number of breathing stops, the highest-risk group still died at more than twice the rate.
Age, sex and weight do much of the work. At Stanford, a similar AI looked at pairs of patients where one died first and picked the right one 84 times out of 100. A simple model using only age, sex, weight and race got 79.
Both studies looked at people referred to sleep clinics, and neither tested a smartwatch or ring.
5
6
49
3,654
Avi Roy retweeted
The wait is over.
Temple Batch 01 pre orders open November 1. Each @temple is hand-made to order and delivered to you within three months of your purchase.
Temple Batch 01 pre orders open November 1. Each @temple is hand-made to order and delivered to you within three months of your purchase.
67
47
1,252
66,989