Drug–Diagnostic Co-Development: Why the Test Runs Behind the Drug (and How to Fix It)

Drug–Diagnostic Co-Development: Why the Test Runs Behind the Drug (and How to Fix It)

Ask anyone who has worked on a biomarker-selected oncology programme where the schedule pressure sat, and many will point to the diagnostic. The drug team has its Phase 3 readout and its filing date. The companion diagnostic (CDx) team is often still locking the assay, chasing samples for a bridging study or waiting on a notified body.

The science is rarely the problem. The sequencing is. CDx development tends to start later than the drug, depends on decisions the drug team has not yet made, and has to pass through a different set of regulators on a different clock.

This guide explains why the test so often lags the drug, what a realistic drug–diagnostic co-development timeline looks like from Phase 1 to approval, and how pharma and diagnostics partners can structure the work so that both arrive together.

Key takeaways

• FDA’s stated expectation is that a CDx and its drug are approved contemporaneously in most circumstances. When the test lags, the cause is usually in the programme plan rather than the review itself.
• The six most common causes of the lag: late biomarker decisions, clinical trial assay to final CDx bridging, sample access, IDE requirements, regulatory sequencing across FDA and IVDR, and partner contracting.
• Under the EU IVDR, a CDx is class C and needs a notified body assessment plus a consultation with EMA or a national medicines authority, which can take up to 120 days on its own.
• Closing the gap means making biomarker decisions at Phase 1, selecting the CDx partner before the pivotal trial, running analytical validation in parallel and designing the bridging study before the first patient is enrolled.
• The partnership contract (fee-for-service or risk-sharing, milestones, IP and data rights) decides how fast the diagnostic can move when the drug plan changes.

What “co-development” actually requires

A companion diagnostic is an in vitro diagnostic that is essential for the safe and effective use of a specific medicine. The EU IVDR uses almost exactly that wording in its definition (Article 2(7)), and FDA’s framework works the same way: the drug label refers to the test, and the test’s intended use refers to the drug.

In practice this makes the drug and the diagnostic co-dependent assets. Neither can launch properly without the other, and your IVD partner is not simply a testing vendor. They are a regulatory co-developer, with their own submissions, quality system, manufacturing and commercial obligations.

FDA set out its position in its 2014 final guidance, In Vitro Companion Diagnostic Devices: in most circumstances, a CDx and its corresponding therapeutic should be approved or cleared contemporaneously. The exception is a new drug for a serious or life-threatening condition with no satisfactory alternative, where the benefits are so pronounced that they outweigh the risk of having no approved test. FDA’s 2016 draft guidance, Principles for Codevelopment of an In Vitro Companion Diagnostic Device with a Therapeutic Product, repeats that expectation and describes how the two programmes should be aligned. For the step-by-step PMA process, see our guide to FDA companion diagnostics and PMA.

Does the test really lag the drug? What the evidence shows

There is no single, well-sourced figure for how far CDx development trails drug development, and we would treat any precise number with caution. What the public record does show is two things.

First, at the point of approval, the first CDx usually arrives with the drug. A 2026 ISPOR poster analysing 185 FDA CDx approvals from 1998 to 2024 found a median lag of about zero years between drug approval and the first CDx, but a median of 2 to 4 years for a second CDx for the same drug and biomarker (Shajarizadeh and Fazeli, ISPOR 2026).

Second, when the diagnostic programme falls behind, the gap can be measured in years. Two well-documented FDA examples:

Contrast these with programmes where both arrived together. In August 2011 FDA approved vemurafenib (Zelboraf) with the cobas 4800 BRAF V600 Mutation Test, and nine days later crizotinib (Xalkori) with the Vysis ALK Break Apart FISH Probe Kit, both ahead of their goal dates. In September 2020, FDA’s accelerated approval of pralsetinib (Gavreto) for RET fusion-positive NSCLC came with the same-day approval of the Oncomine Dx Target Test as its CDx.

So the real problem is less about approval dates and more about what happens inside the programme. When the test runs behind, the drug team either waits, launches with laboratory developed tests (LDTs) and a post-marketing commitment, or compresses CDx validation into the months before filing. FDA’s 2023 Oncology Drug Products Used with Certain In Vitro Diagnostics Pilot Program exists precisely because some oncology drugs are approved before a CDx is authorised, leaving LDTs to select patients.

Why CDx development falls behind: six causes

1. Late biomarker decisions

The diagnostic cannot be locked until the drug team commits to the biomarker, the specimen type, the scoring method and the cut-off. In many programmes those decisions slip into Phase 2 or later, because the team wants more data before narrowing the population. Every month of hesitation upstream is a month the CDx partner cannot spend on design lock and analytical validation.

2. Clinical trial assay vs final CDx: the bridging study

Early trials often enrol patients using a clinical trial assay (CTA), a central lab LDT or local testing. If the final commercial CDx differs from the assay used to select trial patients, the sponsor has to show concordance between them in a bridging study, usually by retesting stored trial samples. The larotrectinib example above shows how this works in practice: the CDx was validated after approval, on archived tissue from trials that had used several local methods.

3. Sample access

Bridging and clinical validation depend on having enough residual tissue or plasma, collected and stored under suitable conditions, from both biomarker-positive and biomarker-negative patients, with consent that covers future diagnostic testing. Archived samples collected without CDx-grade protocols rarely meet the standard needed. If the protocol only screened in positives, or tissue blocks were exhausted, the evaluable population shrinks.

4. IDE requirements

In the US, an investigational IVD used to select patients or assign treatment in a drug trial may be a significant risk device, which requires an investigational device exemption (IDE) under 21 CFR 812 as well as the drug IND. Since FDA finalised its streamlined submission guidance for investigational IVDs in oncology trials in October 2019, sponsors can include IVD information in a new IND and receive a study risk determination within the 30-day IND review. But a significant risk finding still means an IDE, and the streamlined route does not apply to amendments to existing INDs. Discovering the IDE question late can delay first patient in.

In the EU, IVDR Article 58(2) subjects performance studies involving companion diagnostics to the same requirements as interventional performance studies, unless they use only left-over samples (which must still be notified to the competent authority).

5. Regulatory sequencing: FDA vs IVDR

In the US, the drug (CDER or CBER) and the CDx PMA (CDRH) are reviewed in coordination, aiming at contemporaneous approval. In the EU the path is split. The medicine goes through EMA or national procedures; the CDx is a class C device under IVDR Annex VIII Rule 3 and needs a notified body to assess its technical documentation (Annex IX, Section 5.2). Before issuing a certificate, the notified body must seek a scientific opinion on the device’s suitability for the medicine from EMA (for centrally authorised medicines) or a national medicines authority. Under Annex IX of the IVDR, that authority has 60 days from receipt of all the necessary documentation to give its opinion, and the period can be extended once for a further 60 days on justified grounds. That time sits on top of the notified body’s own review and any questions. Our IVDR companion diagnostics guide covers the process in detail.

6. Partner selection and contracting delays

Selecting a CDx partner involves platform choice, global footprint, regulatory track record and capacity. The contract then has to settle development costs, milestones, IP and data rights, regulatory responsibilities and commercial obligations. These negotiations can run for months, and the clock on analytical validation does not start until they are done.

Drug vs CDx workstreams: a timeline from Phase 1 to approval

The table below shows how the two workstreams should run in parallel. The right-hand column shows where programmes most often fall out of step.

How to close the gap

Set the biomarker strategy at Phase 1

Treat the biomarker as a development asset from the first-in-human study. Define the hypothesis, the likely specimen type and the candidate technologies early, and bank samples from Phase 1 onwards with consent that covers future diagnostic use. Even if the final cut-off is not known, you can narrow the platform choice and protect future bridging work.

Choose the CDx partner before the pivotal trial

The best time to sign a CDx partner is during Phase 1b or Phase 2, once there is a credible biomarker signal but before the pivotal trial design is fixed. That gives the partner time to lock the assay so the final CDx, or something very close to it, selects patients for Phase 3. Evaluate partners on platform fit, existing approvals and regulatory track record in the US and EU, manufacturing capacity and installed base in the markets you plan to launch in.

Run analytical validation in parallel

Analytical validation (precision, reproducibility, limit of detection, specimen stability, inter-reader agreement for IHC) does not need to wait for the pivotal readout. Starting it as soon as the assay design is locked takes it off the critical path. Our guide to validating an IO biomarker in a clinical trial covers the analytical and clinical validation steps in more depth.

Design the bridging study before the first patient

If a CTA or LDT will be used for enrolment, write the bridging plan into the pivotal protocol. Retain samples from all screened patients, including biomarker-negatives, specify storage conditions, and agree the statistical approach for concordance and for handling missing results. Discuss the plan with FDA through the pre-submission programme and, where the EU matters, with the notified body early.

Plan for global regulators from the start

Map the FDA PMA timeline and the IVDR route side by side. Identify the notified body early, confirm whether the EMA or a national authority will be consulted, and build the consultation period into the EU launch plan. Make sure the drug’s regulatory team and the CDx team share one integrated plan, so a change to the drug’s indication or timing is reflected in the diagnostic plan immediately. Where the biomarker is already measured by several assays (PD-L1 is the classic example, see our PD-L1 biomarker article), agree early on how cut-offs and scoring will be described in both labels.

Get the commercial and lab network ready

An approved test that is not available in the labs where patients are treated delays access just as surely as a missing approval. Plan for kit supply, lab onboarding, pathologist training, proficiency testing and test reimbursement alongside drug market access. Health economic evidence for the test should be built during development, not after approval.

Structuring diagnostic development partnerships between pharma and diagnostics companies

Most CDx partnership problems are structural rather than scientific. The contract determines how quickly the diagnostic can adapt when the drug plan changes, and who pays when it does.

Fee-for-service vs risk-sharing

In a fee-for-service model, the pharma sponsor pays for assay development, validation and regulatory work, usually against a defined scope and budget. It gives the sponsor control, but scope changes (a new cut-off, an added indication, a switch from tissue to plasma) trigger renegotiation. In a risk-sharing model, the diagnostics partner funds part of development in exchange for downstream value, such as commercial rights to the test, a share of testing volumes or the ability to add the claim to a wider platform. It aligns incentives but needs clear rules on what happens if the drug fails or the indication changes.

Milestones

Milestones are typically linked to development events on both sides: assay design lock, completion of analytical validation, IDE or study risk determination, pivotal trial enrolment, regulatory submission and approval. Linking diagnostic milestones to drug programme events, rather than fixed dates, keeps the two workstreams in step when timelines move.

IP and data ownership

Assay design, biomarker data, software and algorithms, validation evidence and the clinical samples themselves all raise questions of ownership and use. The contract should define who owns the assay and any improvements, who can use trial-derived biomarker data and for what, whether the diagnostics partner can use the validation data to support other claims on the same platform, and how the samples may be used after the study. Without clear definitions, disputes tend to surface at the worst possible moment, during pivotal development or filing.

Governance

The biomarker landscape is making this harder, not easier. Programmes increasingly combine genomic, transcriptomic, proteomic and spatial data, and our review of immuno-oncology biomarker trends shows how many of these approaches are moving towards clinical use. Turning that complexity into a standardised, reproducible, regulatory-grade CDx is exactly where early partnering pays off.

FAQs

What is CDx development?

CDx development is the process of designing, validating and gaining regulatory approval for a companion diagnostic, an in vitro diagnostic that is essential for the safe and effective use of a specific medicine. It covers biomarker and cut-off selection, assay design, analytical and clinical validation, any bridging study, and regulatory submissions such as an FDA PMA or an IVDR notified body assessment.

Does FDA require a companion diagnostic to be approved at the same time as the drug?

FDA’s 2014 guidance says that in most circumstances a companion diagnostic and its drug should be approved or cleared contemporaneously. The guidance also allows FDA to approve a drug before its CDx when the drug treats a serious or life-threatening condition with no satisfactory alternative treatment and the benefits clearly outweigh the risk of having no approved test. Larotrectinib is one example of a drug approved first: its FDA-approved CDx followed almost two years later.

What is a CDx bridging study?

A bridging study shows that the final companion diagnostic gives results that agree with the assay used to select patients in the pivotal trial. It is needed when the trial used a clinical trial assay, an LDT or local tests rather than the final CDx, and it usually involves retesting stored trial samples from both biomarker-positive and biomarker-negative patients.

How long does the EMA consultation for a companion diagnostic take under IVDR?

Under IVDR Annex IX, the medicines authority consulted by the notified body (EMA or a national competent authority) must give its opinion within 60 days of receiving all the necessary documentation. That period can be extended once by a further 60 days on justified grounds, and it sits on top of the notified body’s own assessment.

Meet the CDx and pharma teams at ImmunoMark London, 3–4 March 2027

Closing the gap between drug and diagnostic depends on the right people talking early: translational and biomarker leads, regulatory teams, CDx developers and central labs. Stream 2 of ImmunoMark Summit London, Biomarkers & Companion Diagnostics, brings that community together, alongside the Next-Gen Immuno-Oncology stream, with keynotes, panels, case studies and pre-arranged 1:1 meetings. The 2027 edition of our immuno-oncology conference in London expects 400+ senior attendees.

Register for ImmunoMark London 2027

CDx developers, central labs and platform companies: meet the pharma teams who are choosing their diagnostics partners now. If your company wants to meet this audience, see the London 2027 sponsorship opportunities.