Notes from Paris: France's Precision-Dx Debate Has Moved On
France isn't arguing about whether precision diagnostics matter anymore. The argument now is how to make them reimbursed, routine, and equitable — and that turns out to be a business-model question as much as a clinical one.
Spend a week in rooms in Paris and you hear the same shift. The science is conceded. The infrastructure largely exists. What everyone is actually wrestling with is the plumbing: who pays, who performs, who interprets, and what happens to a patient when the test is needed before the system has officially decided to fund it.
That tension has a very French shape — call it centralized excellence, decentralized friction. France can design a world-class national framework and still get stuck at the toll booth: reimbursement opacity, closed hospital envelopes, accreditation, sample logistics, and the gap between when a therapy is approved and when its companion test is durably funded.
Here's how I'd force-rank what people are actually talking about, scored the same way as my macro-currents table.
Table 1: The precision-Dx "talk of the town" in France (mid-2026)
| Theme | Score | Comment |
|---|---|---|
| RIHN / LAHN reform → routine reimbursement | 94 | The toll booth everything queues at. Voted years ago, still mid-implementation, with HAS grinding through ~127 molecular acts across ~900 clinical situations. Nothing reaches durable reimbursement except through here — highest structural leverage in the system, maximal crossover with every other row. |
| The companion-diagnostic funding gap | 89 | The cleanest burning-platform data point (Leem 2025): 87 CDx-linked therapies, ~420k biomarker searches/yr, ~205k patients, ~58% still on transitional or partial funding → hospital-level and territorial inequity. The symptom that makes the reform urgent. |
| The drug–test paradox | 82 | France grants early/compassionate drug access faster than it can durably fund the test that unlocks it. A dominant framing in policy rooms — though it's the narrative layer over the two rows above more than a separate workstream. |
| PFMG2025 moving to messy scale-up | 80 | The infrastructure exists (AURAGEN, SeqOIA); the debate shifted from "build it" to "make it routine, fast, and funded." Live tensions: turnaround, diagnostic yield, penetration vs. eligible population — and what the operating model becomes once the "2025" branding expires. |
| Liquid biopsy, indication-by-indication | 76 | HAS's move on a ctDNA NGS panel in advanced/metastatic NSCLC — with accreditation and pre-/post-analytical quality conditions — says France will say yes, but bounded and evidence-gated, one indication at a time. A template, not a green light. |
| AI + genomics | 74 | Maximum mindshare (SFMPP 2025 made it a headline engine), but in the French system still mostly pilot, not reimbursed or embedded. Note the inversion vs. the macro-currents table: globally AI ranks #1 by mindshare; here it's pre-structural. The gap between hype and reimbursable workflow is the story. |
| Rare disease & newborn screening | 68 | Where France's public-health model may show best: PNMR4 (2025–30), SMA moving toward a first genetic newborn screen, genome-in-screening pilots. Narrower stakeholder set and a longer commercial horizon, but the strongest "equity + system integration" proof point. |
Scoring based on ubiquity across French stakeholders (HAS/CNAM, hospitals, reference labs, IVD/pharma, clinicians, patients), persistence, crossover with the other themes, and structural impact on how the market operates — not headline volume alone.
Read top to bottom, the table tells one story: in France, the product is not only the test. The product is the implementation pathway — the assay plus the evidence package, the health-economic logic, the workflow, the accreditation route, the clinical champions, and the path from RIHN/LAHN to routine reimbursement.
The bigger question: what does Europe's next-gen Dx market become?
Now zoom out, because the contrast with the U.S. is what makes this interesting.
2025 was a banner year for U.S. specialty diagnostics — Guardant (GH), Natera (NTRA), GeneDx (WGS), Tempus (TEM), even Grail (GRAL). These names stopped looking like single-test companies and got re-rated as scaled precision-medicine platforms that combine clinical evidence, reimbursement access, ordering workflow, bioinformatics, pharma partnerships, and longitudinal patient data. Natera alone grew revenue ~36% to north of $2.3B. The market rewarded the operating model, not the assay.
So the open question for Europe — and France specifically — is how the same modalities (MRD, liquid biopsy, CGP, rare-disease WGS, pharmacogenomics, advanced pathology, and eventually MCED) actually get implemented here. Do they stay embedded inside academic medical centers and national reference networks? Or does Europe grow a new class of scaled specialty-Dx platforms?
It won't simply copy the U.S. The gravitational pull is different: stronger public systems, national reference labs, tendering dynamics, and tighter reimbursement. The likelier outcome is hybrid — centralized where scale, data, and complexity matter; decentralized where access, equity, and health-system integration win. If a European champion emerges, it comes from one of three places:
- National reference networks commercializing outward — the PFMG/AURAGEN/SeqOIA muscle scaling beyond its mandate.
- A pharma-aligned CDx play that absorbs the evidence-and-reimbursement burden the system can't.
- An AI-native entrant that wins on interpretation and workflow rather than chemistry.
And here's the investment angle worth sitting with: if you believe Europe is simply trailing the U.S. by two or three years on this curve — same modalities, same reimbursement logic, just delayed — then the enablers of the European testing wave look a lot like the U.S. specialty-Dx names did before their re-rating. The catalysts are even legible: RIHN/LAHN conversion, IVDR-driven LDT-to-IVD migration, and indication-by-indication HAS approvals. The thesis isn't "the science will work." It's "the operating model and the reimbursement plumbing will eventually catch up" — and whoever has built for that ahead of time captures the value.
The next wave won't be won by the best assay. It'll be won by whoever figures out who orders, who performs, who interprets, who pays, and who owns the data. In the U.S., complexity favored the players who could absorb all of it. The interesting bet is who does that in Europe — and whether they get the same re-rating when they do.