Unlocking the Full Potential of Precision Medicine: Why Companion Diagnostics Must Start at Discovery

Precision medicine has transformed the way we treat cancer, making it possible to match therapies to the molecular profile of each patient and revolutionize treatment approaches, leading to better outcomes. Recent findings from the European Society for Medical Oncology show that patients with advanced cancer lived 2.6 times longer when treated with targeted therapies compared to standard treatments [1].

For precision medicine to be most efficient, the timing of companion diagnostics (CDx) is crucial. Pharmaceutical companies that are proactive instead of reactive with CDx are more likely to avoid challenges that can often come up when diagnostics are added late in drug development, such as trial delays, regulatory setbacks, and manufacturing inefficiencies. Starting earlier, with diagnostics integrated into the drug discovery process, can help laboratories design smarter trials, streamline alignment with regulatory bodies, and move more efficiently toward delivering therapies to patients.

The Case for Early CDx Integration

Developing diagnostics alongside targeted therapies provides a clearer, deeper understanding of the underlying biology. This early insight helps identify predictive biomarkers that guide patient selection and trial design; this approach produces more focused studies, shorter development timelines, and higher success rates. By contrast, late integration often forces rushed decisions and yields scattered data.

Starting early not only improves the likelihood of clinical success but also helps facilitate regulatory approval. Agencies such as the FDA [2] and EMA [3] increasingly encourage co-development strategies, recognizing that early diagnostic integration can streamline the review process and reduce the risk of post-market complications.

Building Stronger Partnerships from the Start

Early planning fosters more effective collaboration. Aligning drug developers, diagnostic experts, and regulators on a common framework ensures robust evidence and mitigates potential challenges in regulatory review. Regulators expect high‑quality, reliable biomarker data; initiating these efforts early ensures teams have the time required to generate it with the necessary rigor.

There are also clear operational benefits that accrue from this approach. Knowing the right patient group from the start lowers risk and helps use resources wisely. Trials that are biomarker-driven tend to be smaller, more focused, and faster to execute. With clearer go/no-go decisions based on early data, teams can avoid investing time and effort into therapies unlikely to succeed.

This kind of early collaboration is already reshaping how cancer is treated. In breast and lung cancers, for example, companion diagnostics are routinely used to match patients with targeted therapies based on specific biomarkers. Similar approaches are gaining traction in melanoma, ovarian, and colorectal cancers, where molecular profiling helps guide treatment decisions and streamline trial design [4]. These real-world applications show how early diagnostic planning doesn’t just improve teamwork but also accelerates therapy development, reduces risk, and ensures that precision medicine delivers on its promise: the right treatment, for the right patient, at the right time.

Digital Pathology and AI: Enabling Precision from the Start

The integration of companion diagnostics at the discovery stage is increasingly supported by advances in digital pathology and artificial intelligence (AI). These technologies allow researchers to analyze tissue samples with pixel-level precision, uncovering subtle biomarker patterns that conventional methods might miss. For example, AI-powered image analysis can quantify HER2 expression levels more accurately, particularly at low levels of expression, which is critical for therapies like Enhertu (trastuzumab deruxtecan) that target HER2-low and ultra-low expressing tumors [5].

This shift from qualitative to quantitative scoring is not just a technical upgrade, but a strategic advantage. Early diagnostic planning that leverages computational pathology enables more reproducible data, better patient stratification, and ultimately, more successful trials. It also helps laboratories meet regulatory expectations for robust, validated biomarker evidence, reducing the risk of delays or rejections.

Multiplex Assays and Multi-Omic Data: The Future of Patient Selection

As precision medicine evolves, so does the complexity of the data required to guide it.

Multiplex assays - capable of measuring several biomarkers at once - are emerging as increasingly influential for the early CDx development. These assays allow researchers to assess not just the presence of multiple biomarkers, but also the broader tumor microenvironment, including expression patterns, spatial distribution, and co-localization of biomarkers.

A multi-omic view has been shown to be valuable across many cancer types, in particular in cancers with heterogeneous profiles, such as ovarian, breast, or colorectal tumors. Early incorporation of genomic, transcriptomic, and proteomic insights grounds trial design in true biological complexity, strengthening the ability to identify patients most likely to benefit. It is also essential to consider how these cutting-edge multi‑omic assays will function in real clinical environments to ensure feasible, scalable deployment. It offers a more informed, inclusive, and forward‑thinking approach to precision therapeutics.

When designing complex CDx solutions, it is essential to optimize technological performance, but also to consider how these tests will be deployed in real-world laboratory settings. It is important to ensure the technologies are compatible with the labs, taking into consideration operational constraints, staffing, existing instrumentation, regulatory requirements, and cost effectiveness. If a CDx cannot be feasibly adopted by the labs that are expected to implement it, even the most promising therapeutic diagnostic pairing risk failing to reach patients who need it. Evaluating real‑world lab infrastructure early helps align assay design, workflows, and technology choices with what labs can realistically adopt, ensuring therapies are scalable, accessible, and ready for clinical impact.

Infrastructure Matters: Building Lab Readiness for Early CDx

Despite the clear benefits, early CDx integration depends on a strong digital infrastructure. Many laboratories still operate with fragmented systems, making it difficult to share data, validate assays, or scale operations efficiently. Interoperability across imaging platforms, data formats, and analytical tools are essential to unlocking the full potential of early diagnostics.

Investments in digital pathology, cloud-based data management, and standardized formats such as DICOM are helping laboratories become more agile and collaborative. Industry-wide efforts to promote best practices and harmonize workflows are accelerating adoption and improving readiness. As the adoption of AI‑enabled biomarkers accelerates, there is a critical need for harmonized, community‑wide guidance to ensure laboratories implement these tools with consistency and clinical rigor. The European Society for Medical Oncology’s landmark 2025 guidance introduces the ESMO Basic Requirements for AI‑Based Biomarkers in Oncology (EBAI), setting a foundational framework for the field [6]. The goal is to create a diagnostic ecosystem where insights flow freely, allowing a quicker, more accurate and robust decisions throughout the development process.

Workflow Examination

Integrating CDx development activities—from algorithm design to assay optimization to real‑world evaluation—within a cohesive framework can substantially speed progress and elevate precision‑medicine results, enabling a faster, more seamless transition from concept to proof‑of‑concept [7].

By leveraging a fully integrated portfolio where components are optimized to work cohesively together, pharmaceutical companies can minimize complications and risks from one step to the next. This reduces data handoffs, lowers the risk of compatibility obstacles, and accelerates development programs.

Key elements of this model include advanced digital pathology platforms, standardized staining technologies, and robust biomarker libraries to support rapid development of prototype assays. Incorporating beta-site testing with leading clinical and academic partners ensures early validation and practical insights, while centralized image and data management enables secure, scalable workflows.

This end-to-end strategy provides continuity from translational research through clinical development, reducing complexity and enabling faster delivery of targeted therapies. Ultimately, it positions stakeholders to meet the growing demand for precision medicine by improving patient selection and accelerating commercialization in an increasingly competitive therapeutic landscape.

In parallel, AI‑enabled assays introduce new operational considerations that make rigorous workflow controls essential. As laboratories transition from conventional pathology to digital and computational approaches, standardized governance frameworks are needed to ensure consistent model and test performance over time. This includes implementing robust quality control measures, reference standards, and monitoring systems to detect and mitigate model or assay drift once deployed in real‑world settings. Establishing these controls early—during development and real world site evaluation—helps maintain analytical validity, safeguards clinical decision‑making, and supports regulatory expectations for AI‑driven diagnostics.

Beyond Oncology: Expanding the Reach of Precision Medicine

While oncology remains the leading field for companion diagnostics, the principles of early CDx integration are increasingly relevant in other therapeutic areas. In autoimmune diseases, for example, molecular profiling can help identify patients who will benefit from biologics targeting specific inflammatory pathways. In neurology, blood-based biomarkers and imaging diagnostics are opening new doors for personalized treatment of conditions like Alzheimer’s disease and multiple sclerosis [8].

These emerging applications underscore the need for flexible, scalable diagnostic strategies that start early and evolve with the science. As the boundaries of precision medicine expand, our approach to diagnostics must evolve as well, shifting from reactive testing to proactive discovery.

Precision medicine promises to transform healthcare, but its success depends on more than just innovative therapies. It requires a diagnostic mindset that starts at the beginning of the drug development, not the end. By integrating companion diagnostics into the earliest stages of drug development, we can design smarter trials, build stronger partnerships, and deliver better outcomes for patients.

The future of medicine is personalized. Let’s make sure our diagnostics are, too.

References:

1. Hurvitz, S.A., Kim, S.-B., Chung, W.-P., Im, S.-A., Park, Y.H., Hegg, R., Kim, M.-H., Tseng, L.-M., Petry, V., Chung, C.-F., Iwata, H., Hamilton, E., Curigliano, G., Xu, B., Egorov, A., Liu, Y., Cathcart, J., Bako, E., Tecson, K., Verma, S., & Cortés, J. (2024). Trastuzumab deruxtecan versus trastuzumab emtansine in HER2-positive metastatic breast cancer patients with brain metastases from the randomized DESTINY-Breast03 trial. ESMO Open, 9(5), Article 102924. Retrieved from https://www.esmoopen.com/article/S2059-7029(2401859-3/fulltext)
2. FDA. (2014). In Vitro Companion Diagnostic Devices: Guidance for Industry and Food and Drug Administration Staff. U.S. Department of Health and Human Services, Food and Drug Administration. Retrieved from https://www.fda.gov/media/81309/download
3. European Medicines Agency. (2023). Frequently asked questions on medicinal products development and assessment involving companion diagnostic (CDx). EMA/CHMP/821321/2022. Retrieved from https://www.ema.europa.eu/en/documents/other/frequently-asked-questions-medicinal-products-development-and-assessment-involving-companion-diagnostic-cdx_en.pdf [www.ema.europa.eu]
4. Ferrari V, Mograbi B, Gal J, Milano G. Companion Tests and Personalized Cancer Therapy: Reaching a Glass Ceiling. Int J Mol Sci. 2024 Sep 17;25(18):9991. doi: 10.3390/ijms25189991. PMID: 39337479; PMCID: PMC11431990. Retrieved from https://pmc.ncbi.nlm.nih.gov/articles/PMC11431990/
5. Aidt F, Arbel E, Remer I, Ben-David O, Ben-Dor A, Rabkin D, Hoff K, Salomon K, Aviel-Ronen S, Nielsen G, Mollerup J, Jacobsen L, Tsalenko A. Quantification of HER2-low and ultra-low expression in breast cancer specimens by quantitative IHC and artificial intelligence. J Pathol Inform. 2025 Aug 26;19:100513. doi: 10.1016/j.jpi.2025.100513. PMID: 41049354; PMCID: PMC12495233. Retrieved from https://pmc.ncbi.nlm.nih.gov/articles/PMC12495233/
6. ESMO Basic Requirements for AI-based Biomarkers In Oncology (EBAI). DOI: 10.1016/j.annonc.2025.11.009. Retrieved from https://www.esmo.org/society-updates/esmo-publishes-first-ever-guidance-on-validation-requirements-for-ai-based-biomarkers-in-oncology
7. Marques L, Costa B, Pereira M, Silva A, Santos J, Saldanha L, Silva I, Magalhães P, Schmidt S, Vale N. Advancing Precision Medicine: A Review of Innovative In Silico Approaches for Drug Development, Clinical Pharmacology and Personalized Healthcare. Pharmaceutics. 2024 Feb 27;16(3):332. doi: 10.3390/pharmaceutics16030332. PMID: 38543226; PMCID: PMC10975777. Retrieved from https://pmc.ncbi.nlm.nih.gov/articles/PMC10975777/
8. Kang SL, Kwon JY, Kim SM. Companion Diagnostics (CDx) Based on Molecular Biology Techniques. Life (Basel). 2024 Oct 23;14(11):1358. doi: 10.3390/life14111358. PMID: 39598157; PMCID: PMC11595734. Retrieved from https://pmc.ncbi.nlm.nih.gov/articles/PMC11595734/