Building Patient-Specific Drug Screens in Weeks, Not Months: A New Model for Precision Oncology

When a patient is diagnosed with cancer, every day counts. For oncologists, the race is on to understand that individual’s disease—what makes it grow, what might stop it, and how to make the right call before time runs out.

That’s what drove Sean Porazinski, PhD, Application Science Lead at Inventia Life Science, to set a bold challenge: go from matrix optimization to drug response data as fast as possible using patient-derived tumoroid models. Not for academic curiosity. But to prove that personalized medicine could move at the pace a patient truly needs.

He wasn’t working with simplified cell lines or off-the-shelf models. These were complex, heterogeneous tumoroids grown from real patient tumors, rich in the molecular quirks and resistance patterns that make cancer so difficult to treat.

And yet, eight weeks later, Sean had results. A functional screening pipeline, personalized drug responses, and a new sense of what’s possible when technology meets urgency.

Building a Model That Reflects the Patient

To develop a meaningful screening pipeline, Sean needed models that truly reflected the tumors they came from, not just genetically but functionally. He turned to two colorectal cancer tumoroid lines from Thermo Fisher Scientific’s OncoPro™ collection. These patient-derived models offered more than just convenience; they preserved the complexity, heterogeneity, and molecular signatures of real tumors—crucial for predicting how a patient might respond to treatment.

But even the most advanced models can fall short without the right environment. That’s where RASTRUM came in.

“Having access to a comprehensive knowledge base on cell–matrix matching made all the difference,” Sean says. “We could tune the matrix to mimic the tumor microenvironment—fibronectin, laminin, collagen I—at the right stiffness to support growth and preserve cellular identity.”

What might take weeks of trial and error in other systems became a single PrintRun with multiple experimental conditions. RASTRUM Allegro allowed Sean to test several matrix compositions and cell densities simultaneously, minimizing cell use and maximizing insight.

“That flexibility lets me work fast, but also smart,” he adds. “We could iterate in real time, which is critical if you’re thinking about how to apply this kind of workflow in a clinically-relevant window.”

A Plug-and-Play Shortcut to Tumoroid Complexity

One of the biggest bottlenecks in personalized medicine is time. Deriving tumoroid cultures from patient biopsies takes weeks (and more often months) and demands specialized expertise that many labs simply don’t have. That’s where the OncoPro™ Tumoroid Cell Lines and Culture Medium come in.

“These models were designed to give researchers a plug-and-play option,” says Colin Paul, PhD, a Cell Biology R&D Scientist at Thermo Fisher Scientific. “They retain the molecular and phenotypic fidelity of the donor tissue, but they’re standardized enough to fit into existing workflows.”

The colorectal tumoroid lines Sean selected had already been carefully characterized, giving him a head start without sacrificing biological complexity. “Traditional cancer cell lines have been foundational,” Colin notes, “but they don’t always reflect what patient tumors look like. OncoPro was built to enable cancer research with physiologically relevant cell models.”

For Sean, the benefit was immediate. With robust growth in 3D and consistent behavior across passages, he didn’t need to troubleshoot viability or spend time coaxing the cells to propagate using complex media formulations. “They just worked,” Sean says. “Once I had the cultures established, I could rely on having enough material for multiple downstream assays with minimal intervention—and that reliability let me focus on optimizing the rest of the system.”

It’s this combination—the accessible, well-characterized biology of OncoPro and the streamlined scalability of RASTRUM—that sets the stage for what’s next. Together, they provide a rapid, reliable foundation for optimizing workflows that could one day be applied directly to patient samples. By removing common barriers like inconsistent cell access or manual handling, this approach frees researchers to focus on what truly matters: running meaningful experiments, generating actionable insights, and ultimately translating those learnings into real-world impact for patients.

Biology That Doesn’t Compromise Under Pressure

Speed is only meaningful if the biology holds up. For Sean, one of the most gratifying moments came when the printed tumoroids began to behave just like their suspension-grown counterparts.

“Each tumoroid line maintained its defining features,” he says. “The morphology we observed in suspension, like cystic structures or differences in growth rates, were still there after printing into the RASTRUM matrices.” For a workflow designed to move quickly, this level of fidelity was anything but guaranteed.

The models didn’t just look right, they grew right, too. Differences in tumor aggressiveness were preserved, with one line forming more irregular, cyst-like structures and the other remaining compact and uniform, in line with its lower-grade clinical origin. “That consistency gave me confidence that what we were seeing in downstream assays—like drug response—was genuinely reflective of the patient biology,” Sean explains.

For Colin, it was a clear demonstration of what these models were designed to do. “We knew the tumoroid models were stable and representative,” he says. “What stood out to me was how quickly the team was able to scale them using the RASTRUM platform—generating 3D models that worked seamlessly across high-throughput drug screening and downstream assays like imaging and immunofluorescence.”

What emerged was a model system that was both scalable and faithful to the patient. No compromises, just real biology captured and carried through an entire workflow.

From Non-Responder to a New Therapeutic Lead

As the drug assays came in, the models revealed a familiar but still sobering pattern: resistance to standard-of-care therapies.

“We saw minimal response to the chemotherapeutics typically used for these molecular subtypes,” Sean explains. “It was disappointing, but also very real. This is the clinical challenge we’re up against—patients who don’t respond to first-line treatments.”

Rather than signal failure, this opened a door. With RNA-seq data already in hand, Sean and the team turned their attention to alternative strategies. One biomarker stood out: LCN2, an immune-related protein overexpressed in both tumoroid lines and in the original patient tissue. Sean’s prior research had linked LCN2 expression with sensitivity to itraconazole—a common antifungal drug increasingly being explored for cancer repurposing.

“It was a natural next step,” Sean says. “We had the biology, the rationale, and a fast, reproducible system to test it.”

The results were striking. Both tumoroid lines responded to itraconazole with IC₅₀ values significantly lower than those reported in traditional 2D CRC cell lines. “It was validating,” Sean adds. “Not just that the models were biologically relevant, but that they could be used predictively to uncover therapies that might otherwise be overlooked.”

Colin echoes that sentiment. “This is exactly the kind of application we envisioned for OncoPro,” he says. “You take a well-characterized, patient-derived model, combine it with a flexible platform like RASTRUM, and you can start answering complex therapeutic questions much more quickly.”

What began as a test of feasibility became a clear demonstration of how biologically faithful models, when paired with the right tools, can generate meaningful, patient-relevant leads in weeks, not months to years.

A New Path Forward

The success of this rapid screening workflow wasn’t just about proving a technical point—it was about imagining a different future for cancer research and care.

“In my academic days, we’d spend hours seeding plates by hand—sometimes more than 20 at a time—for a single drugging screen,” Sean recalls. “It was time-consuming, inconsistent, and frankly exhausting. With RASTRUM Allegro, I could generate enough plates for multiple downstream assays in one experiment, using fewer cells and with complete reproducibility.”

That shift, from manual labor to a streamlined workflow, does more than save time. It reduces variability, increases confidence in the data, and frees scientists to focus on interpreting results rather than troubleshooting workflows. “And when you're working with patient-derived models,” Sean adds, “every bit of consistency matters. These models have inherent heterogeneity, which is valuable, but it also means you need to control every other variable.”

For Colin, the broader implications are clear. “As personalized medicine becomes more mainstream, we need tools that can scale while preserving the complexity of human biology,” he says. “OncoPro and RASTRUM are a natural fit for that challenge—bringing together standardization, flexibility, and fidelity.”

Sean envisions a future where platforms like RASTRUM and OncoPro media could support more direct testing of patient samples, even within hospital or pathology lab settings. “It’s exciting to think about how this kind of workflow could eventually help inform treatment decisions more quickly,” he says.

While still a few steps away, this case study offers a glimpse of what that future could look like.

From Weeks to Insight—Built for the Pace of Patients

In the world of cancer, time isn’t just precious—it’s personal. Every delay in understanding a tumor’s biology or identifying a therapeutic option can mean missed opportunities for intervention. That’s why workflows like this matter.

By combining robust, patient-derived tumoroid lines with the flexibility and reproducibility of RASTRUM Allegro, Sean and the team moved from matrix optimization to actionable drug response data in just eight weeks. No compromises. Just reproducible, high-quality science fast enough to matter.

That eight-week turnaround is more than just a milestone—it aligns closely with national initiatives like Omico’s Cancer Screening Program (CaSP), which typically delivers comprehensive genomic and treatment reports within 8–10 weeks of patient consent. That’s no coincidence: Sean previously collaborated with Omico on their Molecular Screening & Therapeutics (MoST) programme at the Garvan Institute, helping shape what timely precision oncology can look like in pancreatic cancer.

While the workflow here used established tumoroid lines, it could be adapted to grow cancer cells directly from fresh patient biopsy/surgical samples, potentially streamlining future assays to focus on direct printing, drug testing, and therapeutic insight. In this setup, the primary rate-limiting factor would be tumor cell growth itself. Still, the framework is flexible and broadly applicable across cancer types that can be modeled using RASTRUM—offering a powerful vision for how personalized treatment decisions might one day be made not just faster, but more accurate.

Explore the full workflow, data, and methods in our latest application note ➝