On Long Island’s North Shore, scientists at Cold Spring Harbor Laboratory are quietly raising tiny stand-ins for patients’ cancers. These lab-grown tumor copies, known as organoids, let researchers test treatments in plastic dishes instead of on people whose time and strength are already in short supply.
The goal is straightforward, even if the science is not: use these microscopic “avatars” to run through many drug combinations at once, then point patients toward regimens that hit their tumors harder and come with fewer side effects. The approach has moved beyond the “cool idea in a petri dish” stage and into early real-world trials, although making it fast enough to help doctors in live cases remains the biggest hurdle. Local hospital systems have already begun enrolling patients in studies that pair tumor biopsies with organoid testing.
“The notion of an organoid is an avatar. It's a mimic of that patient's tumor you grow in a laboratory dish,” Dr. David Tuveson, director of the Cold Spring Harbor Cancer Center, told reporters. As reported by Spectrum News, Tuveson’s group is starting with pancreatic cancer, one of oncology’s toughest cases, and current organoid production can still take weeks to months.
Clinicians such as gastrointestinal oncologist Dr. Daniel King at Northwell Health are now offering eligible patients the option to have their tumors modeled on these platforms as part of clinical trials. Trial records indicate that organoid samples are being collected alongside other biomarkers so researchers can compare how drugs perform in the dish with how those same treatments perform in patients, according to ClinicalTrials.gov.
High-throughput testing in a dish
“Not only can we study one drug at a time, we can combine multiple drugs at a time in a dish,” King told reporters, explaining why organoids are so tempting for oncologists who want to test treatment cocktails without gambling on a person’s only shot, as reported by Spectrum News.
In practical terms, labs seed hundreds of these mini-tumors onto trays and expose them to different drugs and combinations, then watch to see which regimens make the cancer cells “unhappy.” Researchers at Cold Spring Harbor and partner institutions are working to standardize those tests so that results from one center can be reliably compared with another and relayed back to treating physicians in a form they can actually use.
Early evidence and limits
There are early hints that the strategy might work as advertised. A recent cohort analysis in the Journal of Clinical Oncology found that patient-derived organoid assays could predict overall response rates and progression-free survival for groups of pancreatic cancer patients, suggesting that behavior in the dish can mirror what happens in the clinic.
But organoids are not magic crystal balls. Reviews and methods papers in Clinical Cancer Research and Molecular Biomedicine point out that these cultures leave out key parts of a real tumor’s neighborhood, including immune cells, blood vessels and supportive stroma. On top of that, not every patient’s cancer will successfully grow into an organoid line, which keeps the technique from being applied universally, at least for now.
What researchers are doing next
To push organoids from bench to bedside, Cold Spring Harbor’s Lustgarten-funded lab is scaling up, training collaborators and running coordinated trials that pair organoid pharmacotyping with patient care, according to the Lustgarten Foundation and Cold Spring Harbor Laboratory. Teams are also trying to automate the finicky culture process and cut assay times from weeks down to days so that results actually arrive within a standard treatment decision window.
For patients and clinicians, the promise is straightforward: if an organoid-derived readout can reliably flag what will and will not work, it could spare people months of ineffective therapy and needless toxicity. Experts stress that the approach still needs broader validation. But as trials proceed and labs refine faster-growing systems, researchers say the way oncologists choose treatments could shift from educated guesswork to data-driven decisions prepared in a dish.
