Beyond the Blueprint | The Next Frontier of Precision Oncology

Session Focus

Ting framed the arc. The first era of precision oncology was largely about identifying the genetic alterations driving a tumor. The panel addressed what comes after that: drugging targets long considered undruggable, the complexity of biomarkers for newer modalities, biology beyond the tumor itself, and the economics that determine whether any of it reaches patients.

Why Undruggable Targets Became Druggable

Rahul Aras PhD, whose company is developing a drug against Wnt beta-catenin, said the timing of the panel was good. Pathways like his have been understood for 30 or 40 years, with no question about their role in cancer. What changed is not the pathway but the approach.

He identified two shifts.

The first is sharper biology. Earlier inhibitors produced toxicity because they hit pools of beta-catenin that normal biology requires. The cancer-causing biology is transcription, and it is now possible to target that specifically. His company degrades nuclear beta-catenin without inhibiting cytoplasmic or membrane-bound pools, which produces anti-tumor activity without the toxicity.

The second is new modalities that did not exist 20 or 30 years ago, making it possible to drug interfaces previously considered too flat to target. He credited several companies working along these lines, noting that in the last six months multiple groups have now shown the pathway can be drugged, including data in hepatocellular carcinoma and in beta-catenin-driven desmoid tumors.

He added a third enabler on the diagnostic side. Herceptin is nearly 30 years old and required immunohistochemistry on tissue. Patient selection now runs through circulating tumor DNA, with further biomarker investigation to understand responders and non-responders, which lets programs move through the treatment landscape considerably faster.

The Investor’s Double-Edged Sword

Adam Kaye, who leads later-stage investing at RA Capital, described what has to be true before the firm invests.

The target needs human validation, genetic or clinical, showing proven activity. There has to be a market need, not only a clinical unmet need today but one that will persist three, five or seven years out. He noted the firm employs 50 MDs and PhDs who spend their time mapping markets for exactly this reason. And there has to be an identifiable patient population today in which to run trials.

He was direct that precision oncology cuts both ways. The upside is right drug, right patient, faster to market, better outcome. The downside is economic: the patient population shrinks and the market opportunity shrinks with it.

He named a practical constraint that compounds this. The market does not test enough. After years of available genetic markers, roughly half of patients are tested, and the panel was convened to discuss expansion from single mutations into something considerably more complex than current testing supports. A trial then has to find those patients and convince them to enroll.

And at the end, the drug still has to be attractive to pharma, which he described as the economic engine fueling the whole system, since an acquisition is what allows the cycle to start again.

Asked whether narrowing the population is worth the timeline benefit, he clarified the strategy he actually wants. Start from a validated target, use precision to identify patients and reach market quickly in a narrow indication, then use the resulting cash flow to expand into multiple other indications.

Biomarkers for Complex Modalities

Diarmuid Moran, who leads precision medicine and diagnostics at Astellas, contrasted the current situation with the field’s origins.

Early precision medicine was comparatively straightforward because the relationships were direct. An EGFR mutation predicted response to an EGFR inhibitor. An ALK rearrangement predicted response to an ALK inhibitor. The hypothesis was obvious.

Immunotherapy illustrates how much harder this has become. The relevant factors extend well beyond the tumor to the interaction between tumor and immune system, the fitness of immune cells, their activation state, and the inhibitory immune cells present.

Layered on top is timing. The system is dynamic, so when a marker is measured becomes as important as what is measured. Translating that complexity into something that can function as a diagnostic is the central difficulty.

He noted the same problem in antibody-drug conjugates, where a clean expression-response relationship might be expected but is not observed, because additional layers of mechanism complicate the biology.

Asked how to balance longitudinal measurement against biomarker robustness, he acknowledged there is no easy answer. Robustness is a required criterion and has to be clinically validated within studies. Translational plans therefore need longitudinal measurement at multiple time points. But the practical question follows immediately: can a test actually be administered as required? Telling a patient a sample is needed a month before treatment may not be workable.

Looking Beyond the Tumor

Sandra McAllister PhD, who runs a translational cancer biology lab at the Mass General Brigham Cancer Institute, took up the biology the field may be missing by focusing on the tumor alone.

Her work centers on age and the aging process, which she has found has substantial effects on how cancers behave and respond. She welcomed the panel’s broader framing on the grounds that important information is being lost when analysis stops at the tumor.

Pricing and the Payer Response

Kaye raised the economics nobody enjoys discussing. Segmenting a disease into a dozen orphan subcategories supports higher prices, which is currently the only way the math works for a narrowed population.

His expectation is that this does not hold indefinitely. At some point payers will zoom out, observe that lung cancer costs three times what it did a decade ago, and push back on the aggregate rather than the individual approval.

Toward Prevention

The conversation closed on whether minimal residual disease monitoring, now used after surgery to detect recurrence, could move earlier into at-risk populations, shifting precision oncology from treatment selection toward interception.

Key Takeaways

1. Undruggable was a statement about approach, not biology. Targeting transcription rather than total protein pools removed the toxicity barrier.

2. Circulating tumor DNA accelerated patient selection. Programs move faster than the tissue-based era allowed.

3. Precision narrows the market as it improves the odds. The strategy is a fast narrow approval funding later expansion.

4. Testing rates are the hidden constraint. Roughly half of patients are tested, against biomarkers growing more complex.

5. Timing is now part of the biomarker. Dynamic systems require longitudinal measurement that must still be clinically practical.

6. Age may be missing from the analysis. Focusing on the tumor alone omits host biology that affects response.

7. Orphan segmentation supports pricing today. Panelists expect payers eventually to respond to the aggregate cost.