Thursday, July 30, 2026 Evidence-led medical research news
The First Cohort
Medical research, with the receipts
LM

Lena Marshall

Clinical trials · FDA · Oncology

A veteran health journalist who reads the trial before the press release. Ten years covering drug development, regulatory decisions, and oncology — now writing independently for The First Cohort.

Drugs & FDA

The FDA Approved the First Cholesterol Pill of Its Kind Before Its Trial Could Show It Prevents Heart Attacks or Deaths

Lipfendra lowers LDL cholesterol by up to 60 percent. The 14,550-patient trial that would show whether that prevents heart attacks or deaths does not finish until November 2029, and Merck is charging roughly $3,800 a year in the meantime.

Cancer

The Same Cancer Mutation, Opposite Effects: What 581 Mouse Tumors Say About Who Gets Sick

Cambridge and Yale gave four strains of mice one identical dose of a carcinogen. The tumors converged on the same pathway but took four different genetic routes to get there, and today's screening ignores the reason.

Investigations

The American Diabetes Association Had Police Remove the Editor of Its Own Journal. Then It Refused to Print His Reply.

Steven Kahn edits the ADA's own journal. In June its meeting had state police walk him out for handing out one of its editorials. In July it refused to print the response he and his colleagues wrote. He calls it what it is: pure censorship.

Nutrition & Food

Pepsi Paid $1.95 Billion for Poppi, in a Category That Averages as Much Sugar as Fiber

UNC researchers read the labels on 108 prebiotic sodas and found roughly 5 grams of fiber, 5 grams of sugar, and, in most cans, a non-nutritive sweetener. Poppi agreed to an $8.9 million settlement over its gut-health claims; months later Pepsi paid $1.95 billion for it.

Cancer

The Evolutionary Fix for Drug Resistance Works Best on a Tumor Too Small to Find

City St George's mathematicians say hitting a tumor at its weakest moment can drive it extinct. Their own model clears 95 percent only below 7 million cells, far smaller than any tumor a scan can find.

Nutrition & Food

In the Lab, Coffee Flipped an Anti-Aging Switch. The Caffeine Barely Did.

Stephen Safe's lab reported that coffee's polyphenols, more than its caffeine, switch on a cellular stress-defense receptor. Every experiment ran in a dish, and the anti-aging headline traveled a great deal further than the data.

Drugs & FDA

For Years, a Foreign Factory Could Put the Active Ingredient in Your Medicine and Never Tell the FDA

The FDA led with a paperwork convenience for manufacturers. Buried underneath was the admission that it can't always trace where a drug's active ingredient was made, a blind spot Congress had to order it to close.

Drugs & FDA

A $440,000 Drug Proved This Obesity Was Never About Willpower

Brain-tumor surgery can wreck the hypothalamus and leave hunger permanently switched on. Setmelanotide restored the missing signal in a real placebo-controlled trial, then arrived priced near half a million dollars a year.

Brain & Aging

For Thirty Years the Model Said Blue Cones Drift Out of Sharp Vision. They Were Turning Into Red and Green.

Johns Hopkins researchers grew human retinas in a dish and watched blue cones convert into the red and green cones behind sharp central vision. The three-decade-old migration model was wrong.

Investigations

Nearly Half of America's ER Substance-Use Visits Involve Someone the Justice System Already Booked

Forty-four percent of Americans who go to an emergency room for substance use have a booking in their past. A survey of 139,524 people measured how much the jail and the ER share a population.

Infectious Disease

David Ho's Lab Built an Antibody to Neutralize Nearly Every Strain of HIV. Its First Test in People With the Virus Held for Two Weeks.

Fifty-four volunteers, nine with HIV, and just three with virus still in their blood. The engineered antibody drove their viral load down as much as ninetyfold, then lost the ground within a week as it drained away.

Metabolic & Weight

For Forty Years He Said the Standard Cholesterol Target Was Wrong. A Computer Finally Agreed.

A Northwestern model projects that the little-used apoB test would prevent more heart attacks and strokes than the LDL number your doctor chases. And non-HDL, a number already printed free on your standard lab report, beat LDL too. No trial has tested any of it on a patient.

Brain & Aging

Your Retina Starts Failing Before Any Scan Can See It. Now Someone Wants to Turn That Into a Panel.

Delayed dark adaptation flagged people who would develop early macular degeneration three years later. Now an optometrist has folded that finding into a 'retinal reserve' framework, bloodwork and all, that no trial has tested.

Infectious Disease

The Nose-Spray TB Vaccine in Your Feed This Week Has Only Ever Worked in Mice

Johns Hopkins fused two genes into a nasal spray, added it to a drug cocktail, and cleared tuberculosis faster in mice. DNA vaccines have dazzled mice for thirty years and disappointed humans nearly every time. This one has not been in a person yet.

Brain & Aging

A Michigan Lab Grew a Living Skull Joint That Wouldn't Fuse. In Mice, It Held.

Craniosynostosis surgery reopens a fused infant skull; in some of the hardest cases it fuses right back and the child goes under the knife again. A triphasic scaffold kept the joint open in mice.

Investigations

Australia Switched Off Social Media for a Generation of Kids. Nepal Already Ran That Experiment.

The ban is sold as youth mental-health protection, but its own university researchers call the harm evidence overblown, and the one nation that already banned a platform got ID checks and VPN workarounds, not healthier kids.

Drugs & FDA

The FDA Cleared a Sickle-Cell Cure for Two-Year-Olds in 53 Days. It Starts With Chemo That Can Sterilize Them.

Casgevy is a real advance against a brutal disease. It is also a $2.2 million therapy that requires sterilizing chemotherapy, saw a child die in trials, and reached toddlers on a 53-day review Makary's FDA hands out as a favor.

Brain & Aging

A Juice Company Built a Department Inside a Japanese University. Its Scientists Just Found Vitamin C Guards the Aging Brain.

Kagome, Japan's largest vegetable-juice maker, funded the work, employs two of its authors, and measured the vitamin itself. The gray-matter correlation it reported is real, and small enough to miss.

Cancer

No Gene Editing, No Catheter Into the Skull: A Pediatric Brain-Cancer T-Cell Trial Reported One Complete Response in 33 Young Patients

Children with diffuse pontine glioma rarely live past a year. A first-in-human trial infused 33 young patients with pediatric brain tumors, including DIPG, with T cells trained on three tumor proteins, no genetic engineering, and reported one complete response and three survivors still alive years l

Drugs & FDA

The FDA Office That Vets Gene Therapy Has Run Through Its Entire Top Tier in a Year

The office that decides whether gene therapies reach patients has lost its director, her deputy, and the man brought in to replace them, all inside twelve months.

Cancer

For Twenty Years, a Vitamin B12 Cancer Drug Has Worked in Mice, Dogs, and Rats. It Still Hasn't Reached a Human Trial.

Nitrosylcobalamin crossed the blood-brain barrier and pooled in a rat's brain tumor. This study didn't shrink one, the synergy ran in a petri dish, and the scientist behind the headlines holds the patents.

Drugs & FDA

The FDA Pulled Infigratinib as a Cancer Drug in 2024. It Just Passed Phase 3 in Children With Dwarfism.

The FDA pulled infigratinib as a failed cancer drug in 2024. Now its maker says the same pill grows children with dwarfism faster, on a surrogate no one has yet shown reaches adulthood.

Drugs & FDA

The Standard Treatment for Nerve Damage Was Never Built to Repair a Nerve

Gabapentin is the seventh most prescribed drug in America, handed out for neuropathy it was never shown to reverse. A retired researcher's drug-free protocol asks the question the pharmacy aisle can't: what if you removed the cause instead?

Cancer

Twenty Years of Nothing on Brain Cancer, and the Best New Lead Is a Vitamin Nobody Can Patent

A University of Calgary trial added high-dose niacin to standard glioblastoma treatment and reported 82% of 24 patients progression-free at six months. The design is single-arm and early, but the molecule is a cheap, unpatentable vitamin.

Cancer

Fibrolamellar Carcinoma Hides Immune Cells in Scar Tissue. A Drug Approved in 2008 Pulled Them Back Out.

Fibrolamellar carcinoma mostly kills young adults, and checkpoint immunotherapy bounces off it. Researchers found the tumor herds T cells into fibrous bands, and a drug shelved for stem-cell work since 2008 freed them in patient tissue. No human trial has started.

Investigations

The WHO Invented Osteopenia for Statisticians. Merck Turned It Into 34 Million Patients.

The man who chaired the 1992 WHO meeting says they never imagined osteopenia would be treated as a disease. Three years later Merck launched Fosamax and started putting bone scanners in doctors' offices.

Drugs & FDA

The Immune System's Guardrails Were Too Rare to Study. A Bonn Lab Just Grew Them in a Dish.

A German lab built a way to grow follicular regulatory T cells from precursors in a dish, opening a notoriously hard-to-study brake on antibody production to direct experiment.

Nutrition & Food

They Call It Aging. Sometimes It's a Vitamin Your Doctor Forgot to Check.

New Cornell research explains how B12 deficiency starves muscle of energy, and why the standard test your doctor orders can call the problem aging.

Nutrition & Food

Summer Sun Won't Save You: A Vitamin D Study Funded by the People Who Sell the Spray

A Newcastle screening study found vitamin D deficiency persists through summer in older and ethnic-minority adults in northern Britain. The funder sells the fix.

Infectious Disease

The Bird Flu Was Hiding in the Udder. The Receptor Map Showed Up Two Years Late.

A new University of Pittsburgh study explains why H5N1 infects cow udders instead of lungs. The receptor evidence pointed there in 2024; the federal response did not move as fast.

Investigations

The Knee Procedure That Barely Beats a Convincing Fake

A university press release pitches a minimally invasive knee procedure as a way to delay joint replacement, but the only sham-controlled trials show the fake version works nearly as well.

Investigations

Fauci's Pardon Was Supposed to End This. It's the Reason He Can't Stay Quiet.

Rand Paul subpoenaed Anthony Fauci on June 22 after declassified documents reopened the COVID-origins question, and the preemptive pardon meant to shield him is why he cannot refuse to answer.

Cancer

A Repair Gene Turns Saboteur, and a $4 Billion Drug Class Sees an Opening

A Penn State lab found that too much of the DNA repair gene EXO1 makes ordinary tumors behave like BRCA-mutant ones in a dish, and respond to the same drugs. The science is clean. The market context is not.

Nutrition & Food

WHO Tripled the Foodborne Death Toll. The Food Didn't Get Deadlier.

WHO's new estimate says unsafe food kills 1.5 million people a year, more than triple its 2015 figure, even as it admits the burden has fallen. The jump is an accounting change, not an epidemic.

Cancer

Ivermectin and Cancer: The Review the Horse-Paste Crowd Wasn't Supposed to Get

A 2025 systematic review finds coherent anticancer biology in ivermectin across ten tumor types. The human evidence to match it does not exist. The question of why is what the punchline was designed to bury.

Cancer

The "Secret Reason" Cancer Drugs Stop Working Was Published in February

A Montreal team's antibody against the SLAMF6 brake on T cells made the rounds this week as a fresh discovery. The paper went up in February, the work is in mice, and the antibody has not been given to a person.

Investigations

'A Reality, Not an Emergency': The Lancet Picks Another Political Side

The Lancet's new editorial recasts immigration enforcement as the public-health pathology. The receipts on the journal that published it run from the Daszak letter to the Surgisphere fraud.

Infectious Disease

The cheaper staph antibiotic just won the biggest trial ever run

The largest randomized trial ever run for staph bloodstream infection finds the cheap generic antibiotic cefazolin works as well as the older standard, with less kidney damage and no pharma money behind the answer.

Metabolic & Weight

Sleep apnea's heart damage may run through the gut, not the airway

Researchers at UCSD report that knocking out a single bile-acid receptor abolishes the arterial plaque sleep-apnea conditions would otherwise drive in mice. The mechanism fits the CPAP era's underwhelming trial record better than the airway story ever did.

Drugs & FDA

The FDA Just Approved a Third Over-the-Counter Naloxone. The Maker Once Doubled the Price of the Injectable.

The agency calls Rextovy's OTC approval a win for access. The manufacturer's pricing history, and the market it just walked into, complicate the story.

Nutrition & Food

Calcium and Vitamin D Got Demoted in the BMJ. The Trial Data Are More Stubborn Than the Headline.

A BMJ meta-analysis of 153,902 older adults reports calcium and vitamin D supplements offer little meaningful benefit against fractures and falls, though the combined arm showed a statistically significant 16% reduction in hip fracture risk.

Drugs & FDA

A Fentanyl Vaccine for Mice. An Overdose Decline for Humans.

Scripps Research published a clever chemistry redesign that let mice's immune systems recognize an entire class of fentanyl analogues. The headlines called it an overdose blocker. The mice are doing fine. Humans are years away, and the overdose decline is already happening without it.

Cancer

A Korean lab found cancer's DNA-repair off-switch. In a mouse.

A Korean state lab reports that a tool compound called UNI418 strips out the DNA-repair proteins cancer cells use to escape PARP inhibitors. The data are preclinical: cell lines and mouse xenografts. UNI418 is not yet a drug.

Infectious Disease

The AI Vaccine That 'Passed' Its First Trial Three Years Ago

Cambridge finished the Phase 1 trial in 2023. The press tour arrived in 2026. The paper, when you read it, said the immunogenicity was modest.

Cancer

ALTAIR Missed Its Primary Endpoint. The Press Tour Found Something Else.

A phase 3 trial testing chemotherapy in ctDNA-positive colorectal cancer patients missed its primary endpoint at p=0.107, and a post-hoc reanalysis is doing the public-facing work.

Cancer

Six melanoma patients, six weeks, one common variable: the question oncology is choosing not to study

An oncologist's bedside pattern, a Senate hearing, and a mechanism paper sitting in a mainstream oncology journal: the case that repeated mRNA boosters may be lifting the brakes off cancers oncologists thought they had quiet is no longer fringe enough to wave away.

Drugs & FDA

The Macaques Get a Reprieve. Pharma Gets a Faster Pipeline.

The FDA's new draft guidance lets oncology developers cut animal studies for certain biologics and conjugated products. The replacement laboratory methods are still being validated.

Cancer

Ozempic's 30 Percent Breast-Cancer Headline, and the Trial Evidence It Walked Past

A single-system observational study at ASCO 2026 attached a roughly 30% lower breast-cancer rate to GLP-1 use. The randomized record already on the table is quieter.

Investigations

The $401 CBC and the $32 CBC Are the Same Test

A Reddit pitch for a Quest reseller put receipts on a markup the insurance-billed lab market has spent years obscuring. A Florida cost study confirms what the consumer market already figured out.

Metabolic & Weight

The Ozempic 'Resistance' Headline the Data Didn't Quite Earn

A new study identifies PAM-gene variants tied to weaker six-month HbA1c response to GLP-1 diabetes drugs, but the 'why Ozempic may not work' framing imports a weight-loss conclusion the paper doesn't reach, and the buried trials that used longer-acting GLP-1 drugs showed no carrier effect at all.

Investigations

A Senate hearing finally put the cancer question on the record. The journal notice had been up for months.

Sen. Ron Johnson's June 3 subcommittee hearing on plausible mechanisms linking mRNA injections to cancer, the South Korean cohort behind it, and the editorial notice Springer had posted on the paper months earlier.

Cancer

Pancreatic cancer's undruggable wall just cracked. The follow-up is 8.5 months and the money was already placed.

Revolution Medicines' daraxonrasib nearly doubled median overall survival versus chemotherapy in previously treated metastatic pancreatic cancer at ASCO 2026, with 8.5 months of follow-up and a two-billion-dollar Royalty Pharma deal already riding on what comes next.

Investigations

Half a Billion in Detainee Health Care, and the Federal Watchdog Quietly Closed

A KFF and AP investigation pulled more than 300 sworn allegations of medical neglect from 33,000 ICE habeas petitions, even as the detention ombudsman office quietly closed and the federal detainee care bill passed half a billion dollars.

Nutrition & Food

Yale's Olive Oil Cancer Story Did Not Actually Test Olive Oil

A Yale mouse study reported in Cancer Discovery is being sold as proof that olive oil fuels pancreatic cancer. The lab fed the mice purified oleic acid, the effect only showed up in males, and no one has tested it in a person.

Drugs & FDA

Andrew Huberman's peptide masterclass and the published evidence

Andrew Huberman's June 1 episode on peptides walks through BPC-157 and seven other compounds. The published human evidence for its lead substance runs to fewer than thirty people.