Cancer
18 stories
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.
CancerThe 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.
CancerPregnancy Stations Killer T Cells in the Milk Ducts, and a Decade Later They Haven't Left
Melbourne researchers found that pregnancy stations killer T cells in the milk ducts, where they persist for at least a decade and track with better breast cancer outcomes. Now the race is on to summon them with a hormone shot instead of a pregnancy.
CancerWhether a T Cell Keeps Killing Cancer or Quits Is Written in How Its Genome Folds
Exhausted CD8 T cells are why checkpoint drugs and CAR-T so often stall. New maps of the 3D genome show their fate is set by how the DNA is physically folded, and by CTCF, the protein that does the folding.
CancerOne Protein Keeps Your T Cells in the Fight. The Same Protein Feeds Autoimmune Disease.
In mice, deleting the LEF1 gene collapsed the reserve T cells that sustain the fight against chronic infection and cancer. The same deletion protected against autoimmune diabetes.
CancerNo 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
CancerFor 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.
CancerTwenty 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.
CancerFibrolamellar 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.
CancerA Cancer Algorithm Found Answers in 71 Children. The 'Quantum' Label Is the Least Interesting Thing About It.
A University of Utah team used the linear algebra of quantum mechanics to pull two survival predictors from just 71 neuroblastoma patients. The science is real. The word doing the marketing is not.
CancerA 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.
CancerIvermectin 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.
CancerThe "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.
CancerA 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.
CancerALTAIR 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.
CancerSix 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.
CancerOzempic'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.
CancerPancreatic 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.