The science is no longer the question
Biomarkers have moved from the margins of drug development into its core architecture.
In 2025, the FDA approved 16 new personalized medicines, approximately 36% of all newly approved therapeutic molecular entities. Personalized medicines have now accounted for at least one third of new US drug approvals for six consecutive years. In the same year, the FDA cleared or approved 16 significant new or expanded companion diagnostic indications.1
Oncology shows the shift most clearly. Of 172 novel oncology drugs approved by the FDA between 2012 and 2025, 95 were precision medicines requiring diagnostic testing for optimal patient selection.5
Exhibit 1
Precision medicine is now a core share of new approvals
Share of FDA approvals that are personalized or precision medicines
The FDA itself describes biomarkers as tools that support medical product development, improve clinical decision making, and enable companion diagnostics tailored to individual patients.2 The scientific direction is settled.
Yet the latest real world evidence points to a less comfortable conclusion. The next implementation challenge is no longer simply whether testing happens. It is what happens after the result becomes available.
Testing is improving, but unevenly
Biomarker testing is now embedded in routine oncology care. The 2026 IQVIA Institute report on US precision oncology shows testing rates for established biomarkers approaching saturation in several tumour types, and next generation sequencing (NGS) used for more than 90% of patients with non-small cell lung cancer (NSCLC).3
Exhibit 2
Testing for established biomarkers is now routine
US testing rates for established biomarkers by tumour type, December 2024 (approximate)
The same data also expose real variation across tumour types, care settings, and methods. Turnaround time is a clear example. Academic centres returned some breast cancer biomarker results one to two days faster than community hospitals. In NSCLC the pattern reversed: community practices returned PCR results six days faster and NGS results two days faster than academic institutions.3
No single setting is consistently ahead. That matters, because days spent waiting for a result are days in which a treatment decision may already be taking shape.
Precision medicine is a chain, not a single decision
A biomarker only creates value if every link in the pathway holds. The right patient must be identified, the right test ordered, a viable sample obtained, an interpretable result produced and delivered on time. Then the clinician must interpret it for the patient and turn it into an appropriate treatment decision.
Progress at one point cannot fully compensate for failure at another. A 95% testing rate means little if the result arrives after therapy has started, or if it is received but not acted upon. The industry has invested heavily in the first five steps. The evidence now suggests the last two deserve equal attention.
The gap is moving downstream
Recent US real world evidence in advanced NSCLC shows that better testing has not, on its own, produced better treatment alignment.
Diaceutics applied a seven step precision medicine framework to data from more than 35,800 newly diagnosed patients, comparing 2019 with 2023.4 Despite improvements in diagnostic testing, the share of patients not receiving the most appropriate treatment barely moved. What changed was where patients were lost.
Exhibit 3
Testing improved, but losses shifted to the treatment decision
Newly diagnosed advanced NSCLC patients, United States, 2019 vs 2023 (n > 35,800)
Losses at the treatment decision stage rose by 14 percentage points. Testing improved, but that improvement did not translate proportionately into treatment aligned with biomarker results and guideline recommendations. Physician behaviour reported in the same research helps explain why.
Exhibit 4
Many treatment decisions are made before the full picture arrives
Share of surveyed US oncologists
These findings are specific to advanced NSCLC in the United States and should not be generalized automatically to every disease, market, or biomarker.
They do, however, reveal a strategically important possibility. As diagnostic access improves, the limiting factor may shift from generating the result to creating the conditions in which the result can reliably influence care.
A biomarker result enters a complex decision system
It would be reductive to call this a physician awareness problem. A biomarker led decision is made inside a system of evidence, diagnostic availability, reimbursement, turnaround times, laboratory capability, guidelines, treatment access, local pathways, multidisciplinary coordination, and patient circumstances.
Earlier IQVIA research identified barriers including inadequate tissue procurement, fragmented continuity of care, inconsistent reimbursement, fragmented testing guidelines, and limited awareness of actionable biomarkers. Its more recent analysis points to three inflection points for biomarker integration:3
- Evidence and alignment: robust predictive evidence, accompanied by regulatory approval and payer alignment.
- Guidelines: inclusion in national clinical guidelines.
- Institutional advocacy: often led by oncologists or pathologists through tumour boards and workflow integration.
This is a more realistic picture of how practice changes. Evidence is necessary, but it works as one part of a clinical, operational, and institutional system.
Even regulatory alignment may not be complete at launch. Of the 95 precision oncology medicines in the Friends of Cancer Research analysis, 20 were approved with a postmarketing commitment related to the analytical and clinical validation of an essential companion diagnostic. Half of those 20 commitments occurred for drugs approved in 2024 and 2025.5
This is not evidence that the system is failing. It shows how fast the science is advancing, and how difficult it is for therapy development, diagnostic validation, regulation, testing infrastructure, and clinical delivery to move at exactly the same speed.
The questions pharma leaders should now ask
For pharmaceutical leaders, the issue is no longer simply whether clinicians know about a biomarker or whether a test is available. The more useful questions are sharper:
- Are complete results available when treatment decisions are actually being made?
- Do clinicians consider the result sufficiently actionable for the patient in front of them?
- Does the report make the clinical implication clear?
- Are guidelines, reimbursement, and local pathways aligned?
- Is uncertainty concentrated around the science, the test, or the treatment decision?
- Which specialties and institutions are successfully translating results into action?
- Where is the pathway breaking down despite apparently strong testing rates?
These distinctions matter because each problem calls for a different intervention.
Exhibit 5
Different breakdowns call for different investments
Where the pathway fails, and where action may need to go
| What the pathway shows | Where investment may need to go |
|---|---|
| Low testing rates | Infrastructure, pathway design, reimbursement, or education |
| Delayed results | Operational redesign of sampling, logistics, and turnaround |
| Results received but not acted upon | Clinical applicability, confidence in the evidence, patient fit, report interpretation, treatment access, or workflow constraints |
Treating all of these as a single "awareness gap" risks directing investment toward the wrong part of the pathway.
From biomarker availability to decision readiness
The biomarker era has arrived. The more consequential question is whether healthcare systems are equally prepared for the decisions those biomarkers create.
The challenge is shifting from proving that precision medicine is scientifically possible to ensuring that its expanding complexity translates into consistent clinical action. That puts physician decision making at the centre, but not in isolation. Physicians are the point where scientific evidence, diagnostic data, guidelines, patient characteristics, institutional pathways, and practical constraints must finally converge.
Biomarkers can reveal the molecular logic of a treatment. They cannot determine whether the result arrives on time, whether it is trusted, whether it is clinically interpretable, whether the therapy is accessible, or whether the clinician considers it right for a particular patient.
That is why biomarkers do not change practice by themselves. They create the possibility of a more precise decision. Practice changes only when that possibility becomes clear, credible, practical, and actionable for the physician responsible for making it.
The strategic question
The next phase of precision medicine may depend on measuring more than test availability and completion. It may require a closer understanding of the interval between receiving a result and acting on it.
Where does a biomarker result gain, or lose, the power to influence a clinical decision?
For senior pharma leaders, this may be the more valuable implementation question. As testing improves, the greatest opportunity may no longer sit only in finding the right biomarker. It may sit in understanding what it takes for that biomarker to become part of the right treatment decision.
References and further reading
- Personalized Medicine Coalition (2026). Personalized Medicine at FDA: The Scope & Significance of Progress in 2025. Read the report
- U.S. Food & Drug Administration. Biomarkers at FDA and Biomarker Guidances and Reference Materials. Biomarkers at FDA Guidances and reference materials
- IQVIA Institute for Human Data Science (2026). Advancing Precision Oncology: Trends in Oncology Biomarker Testing in the U.S. Read the report
- Diaceutics (2026). Clinical Practice Gaps Report. Summary (News-Medical) Coverage (SelectScience)
- Friends of Cancer Research (2025). Companion Diagnostic Postmarketing Commitments for Novel Oncology Drugs Approved from 2012-2025. Read the analysis
© 2026 GNM Healthcare. This article reflects the views of GNM and its contributors and is provided for informational purposes only. It may not be reproduced or redistributed, in whole or in part, without prior written permission.