Testing Drug Responses, Not Just Genes: A New Functional Precision Medicine Trial Comes to Columbia

A national clinical trial now opening at Columbia is testing a novel approach to precision medicine—one that evaluates how a patient’s tumor cells respond to drugs directly, rather than inferring treatment efficacy based on genetic information alone. We spoke with Rob Wechsler-Reya, PhD, who developed the approach and helped launch the trial – focused on childhood brain tumors – through the Pediatric Neuro-Oncology Consortium (PNOC). 

What is functional precision medicine, and how does it differ from traditional precision medicine?

Traditional precision medicine typically relies on tumor sequencing—identifying genetic mutations or changes in gene expression and prescribing drugs thought to target those alterations. Functional precision medicine takes a different approach. Instead of inferring what might work based on changes in RNA or DNA, we directly test how a patient’s tumor cells respond to drugs. 

In our case, we ask a very straightforward question: does a given drug kill a patient’s tumor cells in a dish? By measuring drug responses directly, we’re testing a functional property of the cancer cells rather than making predictions based on DNA or RNA.

How did this approach originate?

We started by doing high-throughput drug screening in the lab using animal models of medulloblastoma. We put tumor cells into multi-well plates, with each well containing the same number of tumor cells but a different drug. After a few days, we measured how many cells were still alive in each well to see which drug was the most effective at killing the tumor cells. 

We then expanded to patient-derived xenograft (PDX) models, which more closely approximate individual patients’ tumor cells. This revealed significant differences in drug response across tumors, which to us, reflected what we see in the clinic – that sometimes, drugs either don’t work, or stop working, even when a genetic mutation may point to efficacy. 

The real turning point came when my colleague John Crawford, a pediatric neuro-oncologist at Rady Children’s Hospital in San Diego, asked whether we could do this on actual patient samples. We tried it and it worked, and that experiment evolved into a routine practice in the pediatric brain tumor clinic.

How did you move this into a national clinical trial?

After demonstrating that the process was feasible—meaning we could get tissue, generate data, convene a tumor board, and make recommendations within a clinically useful timeframe—we proposed the idea to PNOC as a national trial. [PNOC is one of the largest clinical trials consortia for pediatric brain tumors]. We opened the first phase focused on feasibility in relapsed medulloblastoma. The current Phase II trial expands the approach to include both relapsed medulloblastoma and relapsed ependymoma, and is designed to test whether this process improves outcomes compared with historical controls. The trial is now open at multiple sites across the country, including Columbia. 

How does the trial work in practice?

When a patient is enrolled, tumor tissue is collected in the operating room. The sample is sent to a CLIA-certified laboratory at the University of Washington in Seattle, where a high-throughput drug screen is performed. We currently test 231 drugs at multiple doses, so we can determine the potency of each drug. In the Phase I trial, we also incorporated genetic sequencing data, but in the Phase II trial, we made that optional. 

The results are reviewed by a molecular tumor board that includes oncologists, neurologists, cancer researchers and computational biologists from across the country. One key member is a pharmacist with expertise in drug interactions and toxicity, who helps assess the safety of novel drug combinations. The tumor board produces a written report recommending four drugs, including guidance on sequencing and dosing. That report is used by the treating physician and can also help support insurance authorization.

Why was sequencing made optional in the Phase II trial? 

In pediatric cancers, sequencing often doesn’t reveal actionable mutations, and even when it does, it can take several weeks to obtain sequencing results. In contrast, we can get drug screening data within about a week. 

Because sequencing rarely changed recommendations and significantly delayed decision-making, we made it optional when we moved the trial into Phase II. When genetic information is available and relevant, it’s still factored into the tumor board’s recommendations. 

Why hasn’t functional precision medicine been widely adopted before? 

There’s been understandable skepticism about whether drug responses in a dish reliably predict responses in patients. Issues like drug metabolism and the ability of a drug to cross the blood–brain barrier complicate interpretation, especially for brain tumors. 

We address this through careful drug prioritization in the tumor board across a range of experts and by recommending multiple options. Looking ahead, we’re interested in combining this approach with advanced drug delivery strategies, such as convection-enhanced delivery or focused ultrasound, to further improve effectiveness.

Why is this trial significant now? 

There have been very few clinical trials testing functional precision medicine, particularly in pediatric cancers. We don’t know yet whether this approach will improve outcomes—that’s exactly what the trial is designed to answer.

I’m excited that, at the end of this study, we’ll have real evidence about whether directly testing tumor cell responses can help guide treatment decisions for children with relapsed or recurrent brain tumors. If this approach is successful, it could be easily adapted to other cancers as well.