Precision Oncology

Broader Labeling for In Vitro Companion Diagnostics (CDx)

Understanding the FDA's critical paradigm shift towards precision medicine by allowing broader, class-based labeling of companion diagnostic devices to optimize tissue utilization and clinical workflow.

1. The Traditional 1:1 Drug-Test Bottleneck

The dawn of precision oncology was built on the foundation of the In Vitro Companion Diagnostic (CDx). Historically, a CDx was rigidly approved for use with a single, specific therapeutic product. For example, if a pharmaceutical company developed an inhibitor targeting the EGFR exon 19 deletion in non-small cell lung cancer (NSCLC), they co-developed a specific PCR assay to detect that exact mutation. If the patient's tumor tested positive using that specific assay, the CDx dictated eligibility for that specific drug.

However, as precision oncology rapidly evolved, multiple competing drugs within the same class (targeting the identical molecular alteration) entered the market. The rigid 1:1 regulatory framework forced clinical laboratories to purchase, validate, and run different commercial diagnostic kits to prescribe different drugs, even though the underlying biology (e.g., EGFR mutation or ALK rearrangement) was identical. This redundancy caused massive logistical bottlenecks, increased healthcare costs, and most critically, rapidly depleted precious biopsy tissue from patients.

2. The Paradigm Shift to Class-Based Labeling

Recognizing that tying a diagnostic test to a single proprietary drug is inefficient and clinically restrictive, the FDA finalized its groundbreaking guidance on "Broader Labeling for In Vitro Companion Diagnostics." This policy outlines a pathway to expand CDx labeling from a single specific drug to an entire specific group or "class" of oncology therapeutic products.

Under the FDA's broader-labeling framework, a companion diagnostic may be labeled for a defined group of oncology therapeutic products when the evidence supports that use. This is not automatic simply because a test detects the same biomarker: the proposed group and disease indications must be specified, and supporting clinical evidence must be available for each therapeutic product using the same specimen type. An expansion of an already approved label may also require an appropriate regulatory submission and FDA review. A single NGS panel or IHC test can therefore support multiple treatment options when its approved labeling and evidence cover those uses.

3. Implementation Challenges and Clinical Validity

While the regulatory flexibility is highly welcomed, earning broader labeling is not trivial. Diagnostic sponsors must demonstrate robust clinical validity across the proposed therapeutic class. This often involves executing massive retrospective analyses of historical clinical trial data across multiple pharmaceutical sponsors, or generating substantial, high-quality Real-World Evidence (RWE) demonstrating that the diagnostic performs consistently regardless of the specific downstream drug chosen.

The guidance marks a significant paradigm shift in personalized medicine. By streamlining the diagnostic process, the FDA has enabled faster patient access to life-saving precision therapies, conserved irreplaceable biopsy tissue for future testing, and significantly reduced the financial and operational burden on clinical pathology laboratories worldwide.

4. Statistical Framework for CDx Validation Studies

A critical but often underappreciated dimension of CDx validation is the statistical rigor required to support broader labeling claims. Unlike a traditional drug trial where the primary endpoint is efficacy in an all-comers population, a CDx validation study must establish that the biomarker itself is predictive — meaning patients selected by the test genuinely derive differential benefit from the targeted therapy compared to biomarker-negative patients.

When the clinical question is whether a biomarker predicts differential treatment benefit, the analysis should compare the treatment effect across biomarker-defined groups, commonly using a prespecified biomarker-by-treatment interaction in an appropriate statistical model. Comparing outcomes between biomarker-positive and biomarker-negative patients who all received treatment is not, by itself, a valid test of predictive value: the difference may reflect the biomarker's prognostic effect rather than a difference in treatment benefit. The appropriate evidence and analysis depend on the proposed diagnostic claim, therapeutic context, study design, and applicable regulatory requirements.

Furthermore, when a single CDx is being validated across multiple drugs within a class, sponsors must address analytical concordance — the degree to which the test produces consistent, reproducible results across different laboratory platforms, operators, and specimen processing conditions. FDA guidance documents on analytical validation for NGS-based CDx assays, for instance, require extensive limit-of-detection studies, interference testing, and multi-site reproducibility data before class-based claims will be entertained.

5. Real-World Implications: NGS Panels as the Exemplar

One practical example of broader CDx labeling is comprehensive genomic profiling (CGP) panels, including FoundationOne CDx, which has FDA-authorized uses across multiple specified oncology indications. The exact drugs, biomarkers, specimen types, and clinical contexts covered must be checked against the current FDA-approved labeling; a panel is not automatically a companion diagnostic for every therapy or alteration it can detect. Rather than running three separate single-gene PCR assays to determine eligibility for three different EGFR-targeted therapies, a single NGS panel run on a single formalin-fixed paraffin-embedded (FFPE) tumor specimen can provide simultaneous, actionable information for dozens of approved therapies across multiple cancer types.

Using one specimen for a broad molecular panel can reduce tissue consumption and help identify alterations relevant to more than one treatment option. Whether this improves clinical decisions depends on sample quality, assay coverage, turnaround time, availability of an effective therapy, and the test's authorized intended use. A biomarker's use across tumor types also does not mean every assay detecting it is interchangeable: the companion-diagnostic claim must match the specific biomarker, platform, specimen, and therapeutic context.

6. Co-Development Strategy and Industry Implications

The shift to class-based CDx labeling has fundamentally altered the commercial and strategic dynamics between pharmaceutical and in vitro diagnostic (IVD) companies. Historically, a single large pharma company would exclusively partner with a single diagnostics company, creating a tightly coupled co-development agreement where both entities shared risks and revenues. This model, while commercially straightforward, created a fragmented diagnostics landscape where a hospital pathology laboratory could theoretically need five different approved CDx platforms to prescribe five drugs targeting the same oncogenic pathway.

Under the new paradigm, IVD companies are incentivized to seek broader labeling that encompasses multiple therapeutic partners. This creates more complex, multi-sponsor co-development consortia or post-market label expansion strategies where a diagnostics sponsor proactively accumulates clinical evidence across multiple sponsored drug trials to build a portfolio of validated drug-biomarker pairs under a single test label. For pharmaceutical companies, this reduces the mandatory CDx co-development burden, as they may be able to reference an existing broadly-labeled CDx rather than sponsoring an entirely new one, potentially shortening development timelines and reducing regulatory costs.

Broader labeling may allow a single assay to support more than one specified therapeutic use when the evidence and labeling justify it. It does not establish that one comprehensive assay can replace all disease-specific testing, and the future role of liquid biopsy and multi-omic platforms will depend on analytical performance, clinical evidence, and regulatory review.

Toolkit Tip: To assess whether a biomarker is predictive rather than merely prognostic, compare treatment effects between biomarker-defined groups using an appropriate study design and, where suitable, a prespecified biomarker-by-treatment interaction. Comparing Kaplan–Meier curves among biomarker-positive and biomarker-negative patients who all received the same treatment cannot by itself establish predictive value. The Kaplan-Meier Survival Calculator can describe time-to-event data, but it does not independently validate a companion-diagnostic claim.

Companion-Diagnostic Review: Questions to Resolve Early

A broader assay label is not automatic permission to use any test for any therapy. The intended use, evidence, device status, and therapeutic labeling must align.

  • Define intended use: name the drug, biomarker, specimen type, patient population, and clinical decision the test is meant to support.
  • Separate evidence layers: analytical validity asks whether the assay measures the biomarker reliably; clinical validity asks whether the biomarker result is associated with the relevant clinical state or response; clinical utility asks whether using the result supports a meaningful clinical decision.
  • Assess platform and specimen differences: if evidence comes from a different assay, tissue, laboratory workflow, or variant class, document bridging evidence and its limitations rather than assuming interchangeability.
  • Pre-specify the analysis: define the biomarker-positive population, cut-off, handling of indeterminate results, missing samples, and multiplicity before examining outcomes.
  • Coordinate drug and device plans: align the clinical protocol, analytical validation, diagnostic labeling, and therapeutic labeling strategy early with the relevant regulatory teams.

Important limitation

Evidence supporting one assay or biomarker-defined population may not transfer to another. Broader labeling and next-generation sequencing panel use require a case-specific evidence and regulatory assessment; this checklist is educational, not a determination of FDA authorization.