Scott Manalis, PhD
Associate Head, Department of Biological Engineering, Massachusetts Institute of Technology (MIT)
At Dana-Farber Cancer Institute’s Center for Patient-Derived Models (CPDM), researchers are advancing cancer research by recreating patient tumors in the lab. The Center uses two complementary approaches: patient-derived xenograft (PDX) models, which involve implanting tumor samples into specially bred mice that accept human tissue, and patient-derived cell lines (PDCLs), which grow tumor cells in petri dishes, often as 3D organoids or spheroids.
“Unlike traditional lab-grown cell lines, PDX models preserve the complexity of real tumors, making them far more useful in research,” says Keith Ligon, MD, PhD, Founding Director of the Center and Chief of Neuropathology at Brigham and Women’s Hospital/Dana-Farber.
But while PDX models have long been considered the gold standard, PDCLs have emerged as a scalable and cost-effective alternative, offering new opportunities for high-throughput drug screening and mechanistic studies, explains Sonam Bhatia, PhD, Director of the Center. She notes that combining both approaches allows researchers to study tumors from multiple angles and improve predictions about treatment response.
How PDX Models Are Created
The Center has refined its use of PDX models to better capture tumor complexity and improve the predictive power of preclinical research. Creating a PDX model begins with collecting tumor samples during surgery or biopsy and linking them to de-identified clinical data, such as treatment history and outcomes. The tissue is then implanted into immunodeficient mice—special lab animals with deficient immune systems that allow human tumors to grow without rejection.
“This approach allows us to preserve the tumor’s original structure, genetics, and treatment response—something traditional lab methods can’t match,” Ligon explains.
The Next Generation of PDCL Models
Advances in cell culture now allows researchers to grow patient tumors in petri dishes without using mice. These next-generation PDCLs are often grown as 3D organoids or spheroids using techniques that better mimic the tumor’s natural environment.
Unlike traditional cell lines grown on plastic and in undefined serum, these models are cultured on complex matrices resembling the human extracellular matrix and supported by defined, tissue-specific growth factors. Organoids and spheroids offer a cost-effective way to study tumors in high-throughput drug screening and mechanistic research.
Each PDCL undergoes validation to confirm it accurately represents the original tumor. Validated models are biobanked—frozen and stored for future use—and distributed through platforms such as Dana-Farber’s cBioPortal, cancermodels.org, ATCC’s HCMI initiative, and Dana-Farber’s licensing partnerships.
A New Way to Test Drugs: Functional Precision Medicine and Innovative Technologies
“This type of ex vivo testing is now enabled by new technology and has the potential to transform how we perform research and drug development.
Keith Ligon, MD, PhD
The Center is also pioneering a faster, more dynamic approach to drug testing called functional precision medicine (FPM). Instead of relying solely on genetic analysis, this method tests how live tumor cells respond to different drugs in real time.
One approach involves studying acutely grown PDCLs with live optical imaging to observe treatment response over time. This time-lapse imaging generates rich data and improves drug testing efficiency.
Another technique, developed with MIT researcher Scott Manalis, PhD, uses a microfluidic device that analyzes 30,000 tumor cells per hour by measuring each cell’s density—a physical marker linked to cellular health and drug sensitivity. This level of throughput was previously unattainable.
In a study funded by Bristol Myers Squibb, the Center used this device to test tumors surgically removed from PDX-bearing mice. These tumors had already been treated in vivo—that is, within the living animals—allowing researchers to compare known responses with ex vivo predictions, which were made by testing tumor cells outside the body in a controlled lab environment. After isolating individual cells, researchers treated them with various drugs and measured responses within hours and days. They found that cell density was a reliable biomarker for predicting treatment outcomes.
“This type of ex vivo testing is now enabled by new technology and has the potential to transform how we perform research and drug development,” says Ligon. “It’s efficient, scalable, and provides actionable insights far earlier in the process.”
Driving Discovery Across Research Settings
The Center’s models support a wide range of research—from drug development and resistance studies to rare cancer investigations. A 2020 study published in Nature used PDX models and a large number of organoid and spheroid cell lines to map drug responses across cancer types, revealing lineage-specific vulnerabilities and resistance mechanisms. Another study in Cell Reports (2021) leveraged a dataset of lab-grown tumor models to identify biomarkers and potential drug targets, with a focus on rare and hard-to-treat cancers. These studies demonstrate how preclinical models can drive more personalized and effective cancer care.
Advancing Precision Medicine
The Center continues to explore new frontiers in cancer research. One promising direction involves creating matched pairs of tumor models—PDX and organoids—to study the same tumor in different ways. Researchers are also using advanced tools like spatial transcriptomics, which maps where genes are active within a tumor, and single-cell sequencing, which reveals how individual cells behave. These technologies help uncover the hidden dynamics of cancer biology.
In addition, the Center aims to support co-clinical trials, where lab research and patient treatment happen side by side. This approach will help doctors make more informed decisions and improve outcomes in real time.
Next-generation PDX and organoid/spheroid models provide a vital link between the lab and the clinic. By mimicking real tumors more closely than previous methods, they help scientists and clinicians make smarter decisions and develop more effective therapies.
As the Center continues to collaborate with leading research institutions and biopharma companies, these models are poised to play an even greater role in the future of cancer care—supporting faster drug development, reducing failures in clinical trials, and advancing the promise of truly personalized medicine.
Team Members
Sonam Bhatia, PhD
Director, Center for Patient-Derived Models, Dana-Farber Cancer Institute
Keith L. Ligon, MD, PhD
Founding Director of the Center for Patient-Derived Models and Chief of Neuropathology at Brigham and Women’s Hospital/Dana-Farber Cancer Institute
