Precision Medicine in 2026: What the Evidence Actually Shows

Precision medicine has spent two decades as the most-promised idea in healthcare. In 2026 it is finally old enough to be judged on results rather than potential — and the results are more interesting than either its promoters or its critics expected.

Within the past eighteen months, a bespoke gene-editing therapy was designed for a single infant and delivered in months rather than years. The world's largest randomised trial of a multi-cancer blood test reported that it missed its primary endpoint while still shifting late-stage diagnoses. A national biobank crossed half a million linked genomes. A polygenic risk test for eight cardiovascular conditions became orderable by any clinician in the United States. Each of these is a real development. None of them is the uncomplicated triumph that headlines suggested.

This matters now because precision medicine has crossed from the research budget into the operations budget. Clinicians are being asked to interpret probabilistic genomic reports in twelve-minute appointments. Health system executives are being asked to fund sequencing infrastructure whose return is measured over decades. Payers are being asked to reimburse tests whose analytical validity is excellent and whose clinical utility is often unproven. Developers and data scientists are being asked to build interpretation pipelines on reference data that under-represent most of the world's population.

This article is for those professionals. It separates what is established from what is experimental, explains the technical foundations without hand-waving, examines where implementation actually fails, and sets out what a reasonable practitioner should do with the evidence as it stands in August 2026.

---

Executive Summary

• Precision medicine is not one technology. It is a decision framework in which molecular, clinical and environmental data are used to stratify patients. Its evidence base is strongest in oncology, rare disease and pharmacogenomics — and weakest in common complex disease prevention.
• Pharmacogenomics has the strongest general-medicine evidence. The PREPARE study across seven European countries reported roughly a 30% reduction in clinically relevant adverse drug reactions with pre-emptive panel-based genotyping, though its open-label design has been legitimately criticised (Swen et al., 2023).
• Circulating tumour DNA works better for escalation than de-escalation. DYNAMIC showed chemotherapy could be safely reduced in stage II colon cancer; DYNAMIC-III failed to demonstrate non-inferiority for de-escalation in stage III; CIRCULATE, reported at ASCO 2026, suggested benefit from escalation but closed early with limited numbers.
• Multi-cancer early detection is promising but unproven at the population level. The NHS-Galleri trial in 142,000+ participants did not meet its primary endpoint of reducing combined stage III/IV diagnoses, while showing fewer stage IV diagnoses and about a 25% reduction in emergency-presentation cancers.
• Individualised gene editing is now regulatorily real. After the CPS1-deficiency case published in May 2025, the FDA issued draft guidance on a "plausible mechanism framework" on 23 February 2026 (Docket FDA-2026-D-1256).
• Data scale is no longer the bottleneck; representativeness and workflow are. Roughly 86–87% of genome-wide association study participants remain of European ancestry, which materially degrades polygenic score performance in other groups.
• Adoption fails at the last mile. Even in metastatic non-small-cell lung cancer, where testing is guideline-mandated, real-world biomarker testing sits around 82% and shows measurable disparities by race and payer.
• Regulation is converging on data governance, with the European Health Data Space (Regulation (EU) 2025/327) phasing in genomic-data provisions from March 2031.

---

What Precision Medicine Actually Is

Strip away the marketing and precision medicine is a stratification argument. Conventional care treats the average patient in a diagnostic category; precision medicine asks whether measurable biological variation within that category predicts different outcomes reliably enough to justify different actions.

Three words are used loosely and should not be:

• Personalised medicine is the older, broader term, encompassing preference-sensitive and lifestyle-tailored care.
• Precision medicine implies measurement — molecular, imaging, physiological or environmental — used to define subgroups.
• Individualised (or N-of-1) therapy is the extreme case, where the intervention itself is manufactured for one person.

The distinction is not pedantic. It determines the evidence standard that applies. A stratification claim can be tested in a randomised trial. An N-of-1 claim cannot, which is precisely why regulators have had to invent new evidentiary frameworks for it.

A compressed history

The field's conceptual roots predate genomics: blood typing, HLA matching and therapeutic drug monitoring were all precision medicine before the term existed. Three inflection points reshaped it:

1. 2003 — completion of the Human Genome Project made systematic variant discovery possible.
2. 2008–2015 — collapsing sequencing costs turned genomics from a project into a service, while targeted oncology drugs proved that molecular subgroups could define treatment.
3. 2015 onward — national programmes (the US Precision Medicine Initiative, Genomics England, and their successors) shifted the question from discovery to delivery.

The 2015–2026 period is best understood as the field's implementation decade: the science mostly worked; the health systems mostly weren't ready.

---

The Technical Stack

The layers, briefly

Germline genomics identifies inherited variants — diagnostic in rare disease, predictive in hereditary cancer syndromes, and informative for drug metabolism.

Somatic genomics profiles the tumour itself, driving targeted therapy selection and increasingly used for molecular residual disease monitoring via circulating tumour DNA (ctDNA).

Pharmacogenomics links germline variation in genes such as CYP2C19, CYP2D6, DPYD and TPMT to drug metabolism, with dosing guidance curated by the Clinical Pharmacogenetics Implementation Consortium and the Dutch Pharmacogenetics Working Group.

Polygenic risk scores (PRS) aggregate thousands of small-effect common variants into a single continuous risk estimate — statistically powerful at population level, considerably blunter at the individual level.

Multi-omics and phenomics — proteomics, metabolomics, transcriptomics, spatial and single-cell profiling — sit largely in research today, though national cohorts have begun releasing them at scale.

---

Where the Evidence Is Strong, and Where It Is Not

Table 1 — Evidence maturity by application (original OneWise assessment)

Application	Analytical validity	Clinical validity	Clinical utility (outcomes)	Practical status, August 2026
Rare-disease diagnostic sequencing	High	High	Well documented (diagnostic odyssey shortened)	Standard of care in many systems
Somatic tumour profiling for targeted therapy	High	High	Established for specific biomarker–drug pairs	Guideline-recommended; under-delivered in practice
Pharmacogenomic panel testing	High	High	Moderate (one large positive implementation trial, methodological debate)	Reimbursement and workflow limited
ctDNA for molecular residual disease	High	High (prognostic)	Mixed — escalation evidence stronger than de-escalation	Rapidly evolving; not yet universal standard
Multi-cancer early detection (MCED)	High	Moderate	Unproven for mortality; primary endpoint missed in the only population RCT	Investigational for screening use
Polygenic risk scores, common disease	High	Moderate, ancestry-dependent	Largely unproven prospectively	Early clinical availability, limited guideline support
Individualised gene editing	Case-level	Case-level	Single documented cases	Experimental, with a new regulatory pathway

Pharmacogenomics: the quietest success

The PREPARE study genotyped patients across 18 hospitals, nine community health centres and 28 community pharmacies in seven countries before prescribing, and reported roughly a 30% reduction in clinically relevant adverse drug reactions among those with an actionable gene–drug interaction (Swen et al., 2023). Correspondence in The Lancet subsequently questioned whether an open-label design and a gatekeeping analysis could support that conclusion firmly, and the authors defended the design as appropriate for an implementation study. Both positions are reasonable. The honest summary: pre-emptive panel testing is feasible at scale and probably reduces harm, and it has not yet been proven in a blinded trial — which may never be ethically or practically constructible.

Oncology: the field's proving ground

ctDNA offers the clearest illustration of why "the biomarker works" and "acting on the biomarker helps" are different claims. In stage II colon cancer, the DYNAMIC trial used ctDNA to withhold chemotherapy from low-risk patients, roughly halving adjuvant chemotherapy use without a detectable recurrence-free survival penalty (Tie et al., 2022). Extending that logic to stage III, DYNAMIC-III did not meet non-inferiority for ctDNA-guided de-escalation (Tie et al., 2025). At ASCO 2026, the CIRCULATE trial reported improved disease-free survival from ctDNA-guided escalation in mismatch-repair-proficient stage II disease, but closed early for funding reasons, enrolled fewer ctDNA-positive patients than planned and relied on per-protocol analysis.

The pattern is consistent and clinically useful: ctDNA negativity is not yet a safe licence to withhold therapy in higher-risk disease, whereas ctDNA positivity is becoming a defensible reason to intensify it.

Screening: a genuinely instructive disappointment

The NHS-Galleri trial randomised more than 142,000 participants aged 50–77 in England to annual multi-cancer blood testing alongside usual NHS care, with results presented at ASCO on 30 May 2026 and topline data released on 19 February 2026. The trial was specifically testing whether adding a blood test to NHS screening could reduce the combined number of cancers diagnosed at stage 3 or 4 over three years; that primary endpoint was not met, with no overall difference in late-stage diagnoses, although substantially fewer stage IV cancers were diagnosed among those receiving the annual test. Investigators also reported a greater than 20% reduction in stage IV diagnoses in the second and third screening rounds, that just over half of participants with a positive result were diagnosed with cancer, and a 25% reduction in cancers detected in emergency settings.

Two readings are defensible. The sceptical one: a screening test that does not reduce combined late-stage incidence has not yet earned population deployment. The sympathetic one: stage shift and emergency-presentation reduction are meaningful, and mortality — the outcome that ultimately matters — requires longer follow-up. Both readings should be held simultaneously until peer-reviewed publication and extended follow-up arrive.

Rare disease: precision at N=1

In 2025, a team at Children's Hospital of Philadelphia and Penn Medicine designed, manufactured and administered a lipid-nanoparticle-delivered base-editing therapy for an infant with severe carbamoyl-phosphate synthetase 1 deficiency, publishing the case in the New England Journal of Medicine in May 2025 (Musunuru et al., 2025). It remains a single case with limited follow-up. Its significance is procedural rather than statistical: it demonstrated that a bespoke genetic medicine could move from diagnosis to dosing in months.

---

Latest Developments (dated)

• 3 September 2025 — Mass General Brigham and Broad Clinical Labs launched a clinician-ordered polygenic risk test covering eight cardiovascular conditions, developed using genotype and clinical data from 236,393 All of Us participants and externally validated in 53,306 Mass General Brigham Biobank participants. A validation study followed in the Journal of the American College of Cardiology in April 2026.
• October 2025 — Results from PATHFINDER 2, a 35,878-participant MCED study, were reported showing high specificity alongside increased cancer detection; these were single-arm-style performance data, not population outcome data.
• 19 February 2026 — GRAIL released topline NHS-Galleri results. 30 May 2026 — full results presented at ASCO; peer-reviewed publication pending as of August 2026.
• 23 February 2026 — The FDA issued draft guidance, Considerations for the Use of the Plausible Mechanism Framework to Develop Individualized Therapies That Target Specific Genetic Conditions with Known Biological Cause (Docket FDA-2026-D-1256), with comments closing 27 April 2026. It is draft, non-binding guidance — not a new statutory approval pathway.
• March 2026 — Genomics England reported that the Generation Study, which sequences newborn genomes against a panel covering more than 200 childhood-onset conditions, had reached roughly half its 100,000-baby recruitment target.
• June 2026 — The NIH All of Us Research Program issued its largest data release: data from more than 747,000 participants, including over 535,000 whole genomes linked to nearly 482,000 electronic health records, plus first-time proteomics and RNA-sequencing subsets. Trade reporting in January 2026 also described a substantial cut to the programme's budget, which has slowed enrolment.

Read this before citing any of the above

Company press releases, conference abstracts and peer-reviewed papers are three different evidentiary tiers. Several of the 2026 developments above currently sit in the first two. Treat conference-presented results as provisional until full publication.

---

Why Adoption Stalls

Table 2 — Barriers, mechanism and mitigation (original framework)

Barrier	Underlying mechanism	Who it binds most	Practical mitigation
Interpretation capacity	Too few clinical geneticists, genetic counsellors and molecular pathologists	Community and rural settings	Molecular tumour boards; tiered reporting; embedded pharmacist-led PGx services
Reimbursement uncertainty	Payers require clinical utility; trials measure clinical validity	Health systems and diagnostics developers	Utility-focused study design; coverage-with-evidence arrangements
Reference-data bias	GWAS and variant databases over-represent European ancestry	Non-European-ancestry patients	Ancestry-aware scoring; multi-ancestry training data; explicit reporting limitations
Workflow friction	Results arrive after the prescribing decision	Frontline clinicians	Pre-emptive rather than reactive testing; EHR-integrated decision support
Variants of uncertain significance	Incomplete functional annotation	Patients and counsellors	Structured reanalysis policies; functional assay pipelines
Data governance complexity	Cross-border and secondary-use restrictions	Researchers and industry	Federated analysis; early alignment with EHDS timelines

The measurable version of this is instructive. In metastatic non-small-cell lung cancer — perhaps the single best-established precision oncology indication — real-world analysis of tens of thousands of US patients found biomarker testing for at least one marker at roughly 82%, with documented disparities in comprehensive next-generation-sequencing testing by race in earlier cohorts. If the flagship indication leaks a fifth of eligible patients, the constraint is not scientific.

The equity problem is technical, not only ethical

Approximately 86–87% of participants in genome-wide association studies are of European ancestry. Because polygenic scores depend on linkage-disequilibrium structure and allele frequencies that differ across populations, scores trained predominantly on European-ancestry cohorts perform measurably worse elsewhere (Duncan et al., 2019). Deploying such a score uniformly does not merely fail to help under-represented patients — it can systematically misclassify them, potentially widening the disparities precision medicine is often claimed to close. Multi-ancestry cohorts such as All of Us exist substantially to correct this, and correction will take years.

---

Regulation and Governance

Three regulatory currents matter to practitioners:

1. Evidence flexibility for ultra-rare conditions. The FDA's February 2026 plausible-mechanism draft guidance proposes that, where a disease's genetic cause is known and a therapy demonstrably targets it, a single application and master-protocol approach may cover multiple variant-specific products. This is a meaningful loosening — and, as commentary in Health Affairs Forefront has noted, one that raises unresolved questions about manufacturing controls and scope creep toward common diseases.
2. Data governance. The European Health Data Space (Regulation (EU) 2025/327) entered into force on 26 March 2025 and applies in stages, with most secondary-use provisions from 26 March 2029 and human genetic data provisions from 26 March 2031. Organisations building genomic research infrastructure in Europe should be designing to that timetable now.
3. Diagnostic oversight. Laboratory-developed tests, direct-to-consumer genomic products and clinical PRS reports occupy inconsistent regulatory space across jurisdictions. A test being purchasable is not evidence that it is validated for the use a patient assumes.

---

Frequently Misunderstood, and Commonly Repeated

Myth: precision medicine means everyone gets a personalised drug. For the overwhelming majority of patients it means being placed into a better-defined group, not receiving a unique product.

Myth: a polygenic risk score tells you whether you will get a disease. It shifts a probability, usually modestly, and its accuracy depends on ancestry and on the non-genetic risk factors already in the model.

Myth: sequencing costs are the barrier. Interpretation, workflow integration, workforce and reimbursement now dominate the cost and difficulty profile.

Myth: a negative genomic test rules out genetic disease. Coverage gaps, structural variants and unannotated regions all limit negative predictive value.

Mistake to avoid: ordering a test before deciding what result would change management. If no result changes the plan, the test generates anxiety, cost and incidental findings without benefit.

---

Practical Takeaways

For clinicians: Prioritise pre-emptive pharmacogenomic testing for patients starting drugs with established gene–drug guidance; the value is lost when testing is reactive. Before ordering, write down the action each possible result would trigger. Document ancestry-related limitations when communicating polygenic results.

For health system leaders: Fund interpretation capacity in the same business case as sequencing capacity — an underpowered genetic counselling service is the most common single point of failure. Prefer utility-linked pilots over volume targets.

For researchers: Design for clinical utility endpoints, not only validity endpoints, and report performance stratified by genetic ancestry as a default rather than a supplementary table.

For developers and data teams: Treat reference-genome and training-cohort selection as a clinical safety decision. Build reanalysis into the product lifecycle; variant classifications change.

For executives and investors: Model reimbursement, not adoption enthusiasm. The rate-limiting factor for most precision diagnostics is the payer's utility threshold.

For students and generalists: Learn to distinguish analytical validity, clinical validity and clinical utility. Most public confusion about this field collapses into conflating the three.

---

Key Insights

1. Precision medicine's evidence base is deep in narrow places and thin in broad ones.
2. Analytical validity is largely solved; clinical utility is the open question.
3. Pharmacogenomics offers the best near-term return for general medicine.
4. ctDNA currently supports treatment escalation more confidently than de-escalation.
5. A missed primary endpoint in a screening trial is informative, not necessarily fatal.
6. Individualised gene editing has moved from possibility to regulatory design problem.
7. Ancestry bias in reference data is a measurable safety issue, not only an equity one.
8. National biobanks are now large enough that governance and funding stability, not sample size, are the binding constraints.
9. Adoption failures cluster after the laboratory, in interpretation and prescribing workflow.
10. The most valuable professional skill in this field is calibrated scepticism about one's own preferred conclusion.

---

Frequently Asked Questions

What is precision medicine in simple terms?

It is an approach that uses measurable biological, clinical and environmental differences between patients to choose treatment, screening or prevention strategies, rather than applying one protocol to everyone with the same diagnosis.

How is precision medicine different from personalised medicine?

Personalised medicine is the broader, older term. Precision medicine specifically implies measurement-based stratification of patients into subgroups.

Does precision medicine actually improve outcomes?

In some settings, clearly yes — rare-disease diagnosis, several targeted cancer therapies and pharmacogenomic-guided prescribing. In others, particularly common-disease prevention using polygenic scores, prospective outcome evidence is limited.

What is a polygenic risk score?

A single number aggregating the effects of many common genetic variants to estimate relative predisposition to a trait or disease. It is probabilistic, not diagnostic.

Are polygenic risk scores reliable for everyone?

No. Scores derived mainly from European-ancestry data perform less well in other populations, because the underlying genetic architecture differs.

What is pharmacogenomics used for?

Adjusting drug choice or dose based on inherited variation in metabolism and response — for example, thiopurine, fluoropyrimidine, clopidogrel and several psychiatric medications.

Should pharmacogenomic testing happen before prescribing?

Pre-emptive testing captures more value than reactive testing, because results are available at the moment of the prescribing decision. Coverage and workflow support vary considerably by health system.

What is circulating tumour DNA (ctDNA) used for?

Detecting molecular residual disease after surgery, monitoring treatment response and identifying resistance mutations. It is strongly prognostic; whether acting on it improves outcomes depends on the clinical setting.

Did the NHS-Galleri blood test work?

It did not meet its primary endpoint of reducing combined stage III/IV cancer diagnoses. It did show fewer stage IV diagnoses, high positive predictive value for a screening test, and fewer emergency-presentation cancers. Mortality data require longer follow-up.

Is whole-genome sequencing of newborns being introduced?

England is running the Generation Study, a research programme sequencing up to 100,000 newborn genomes against a panel of childhood-onset conditions, with wider ambitions set out in the NHS 10-Year Health Plan. It is research, not established screening policy.

What was the first personalised CRISPR therapy?

A base-editing therapy designed for an infant with carbamoyl-phosphate synthetase 1 deficiency, reported in the New England Journal of Medicine in May 2025.

What is the FDA's plausible mechanism framework?

Draft guidance issued on 23 February 2026 describing how sponsors might generate adequate evidence for individualised therapies targeting genetic conditions when randomised trials are infeasible. It remains draft and non-binding.

Who pays for precision medicine testing?

Coverage varies widely. Diagnostic and companion-diagnostic testing is often reimbursed; predictive polygenic and multi-cancer screening tests are frequently self-pay.

What are the main ethical concerns?

Equity of access and predictive accuracy across populations, informed consent for lifelong data storage, incidental findings, genetic discrimination and psychological impact of probabilistic risk information.

What skills do professionals need to work in this field?

Statistical literacy in risk communication, familiarity with variant interpretation standards, understanding of clinical trial design, and enough bioinformatics to interrogate a pipeline's assumptions.

Is precision medicine cost-effective?

It depends entirely on the application. Targeted testing that averts ineffective expensive therapy can be cost-saving; broad predictive screening in unselected populations frequently is not, and evidence remains contested.

---

Glossary

• Analytical validity — how accurately and reproducibly a test measures what it claims to measure.
• Base editing — a genome-editing technique that chemically converts one DNA base to another without cutting both strands.
• Clinical utility — whether using a test improves patient outcomes.
• Clinical validity — how well a test result correlates with the clinical condition or outcome of interest.
• CPIC — Clinical Pharmacogenetics Implementation Consortium; publishes peer-reviewed gene–drug dosing guidelines.
• ctDNA — circulating tumour DNA; tumour-derived fragments detectable in plasma.
• EHDS — European Health Data Space; the EU framework for primary and secondary use of electronic health data.
• GWAS — genome-wide association study; scans genomes across many individuals to find variants associated with a trait.
• Liquid biopsy — analysis of tumour-derived material from blood rather than tissue.
• MCED — multi-cancer early detection; blood tests seeking signals from many cancer types simultaneously.
• MRD — molecular (or minimal) residual disease; microscopic disease remaining after treatment.
• N-of-1 therapy — an intervention developed for a single patient.
• Polygenic risk score (PRS) — aggregate estimate of genetic predisposition from many common variants.
• Somatic vs germline variants — acquired variants present in tumour tissue versus inherited variants present in all cells.
• VUS — variant of uncertain significance; a genetic change whose clinical impact is not yet established.
• Whole-genome sequencing (WGS) — determination of an organism's near-complete DNA sequence.

---

---

References

Academic Papers

Duncan, L., Shen, H., Gelaye, B., Meijsen, J., Ressler, K., Feldman, M., Peterson, R., & Domingue, B. (2019). Analysis of polygenic risk score usage and performance in diverse human populations. Nature Communications, 10, 3328. https://doi.org/10.1038/s41467-019-11112-0

Musunuru, K., Grandinette, S. A., Wang, X., Hudson, T. R., Briseno, K., Berry, A. M., … Ahrens-Nicklas, R. C. (2025). Patient-specific in vivo gene editing to treat a rare genetic disease. New England Journal of Medicine, 392(22), 2235–2243. https://doi.org/10.1056/NEJMoa2504747

Swanton, C., Johnson, P., Round, T., Warwick, J., Jones, H., Kumar, H., Liang, W., Smittenaar, R., Neal, R. D., & Sasieni, P. (2026). NHS-Galleri: Primary results from a randomised controlled trial to assess the clinical utility of a multi-cancer early detection (MCED) test in population screening. Journal of Clinical Oncology, 44(17_suppl), LBA100. https://doi.org/10.1200/JCO.2026.44.17_suppl.LBA100

Swen, J. J., van der Wouden, C. H., Manson, L. E., Abdullah-Koolmees, H., Blagec, K., Blagus, T., … Guchelaar, H.-J. (2023). A 12-gene pharmacogenetic panel to prevent adverse drug reactions: An open-label, multicentre, controlled, cluster-randomised crossover implementation study. The Lancet, 401(10374), 347–356. https://doi.org/10.1016/S0140-6736(22)01841-4

Tie, J., Cohen, J. D., Lahouel, K., Lo, S. N., Wang, Y., Kosmider, S., … Gibbs, P. (2022). Circulating tumor DNA analysis guiding adjuvant therapy in stage II colon cancer. New England Journal of Medicine. https://doi.org/10.1056/NEJMoa2200075

Tie, J., et al. (2025). Circulating tumor DNA-guided adjuvant therapy in locally advanced colon cancer: The randomized phase 2/3 DYNAMIC-III trial. Nature Medicine. https://doi.org/10.1038/s41591-025-04030-w

The All of Us Research Program Genomics Investigators. (2024). Genomic data in the All of Us Research Program. Nature, 627, 340–346. https://doi.org/10.1038/s41586-023-06957-x

Government and Official Sources

National Institutes of Health. (2026, June 30). NIH's All of Us Research Program is now the largest integrated genomics and health database in the world. nih.gov

U.S. Department of Health and Human Services. (2026, February 23). FDA launches framework for accelerating development of individualized therapies for ultra-rare diseases. hhs.gov

U.S. Food and Drug Administration. (2026). Considerations for the use of the plausible mechanism framework to develop individualized therapies that target specific genetic conditions with known biological cause (Draft guidance; Docket No. FDA-2026-D-1256). fda.gov

Regulations and Standards

Regulation (EU) 2025/327 of the European Parliament and of the Council of 11 February 2025 on the European Health Data Space and amending Directive 2011/24/EU and Regulation (EU) 2024/2847. Official Journal of the European Union, L, 2025/327, 5 March 2025.

Federal Register. (2026, February 25). Considerations for the use of the plausible mechanism framework to develop individualized therapies that target specific genetic conditions with known biological cause; draft guidance for industry; availability (Docket No. FDA-2026-D-1256). federalregister.gov

Institutional and Research Programme Sources

Broad Institute. (2025, September 3). Mass General Brigham launches genetic test to predict risk across eight cardiovascular conditions. broadinstitute.org

Genomics England. (2026). Newborn Genomes Programme / Generation Study. genomicsengland.co.uk

Innovative Genomics Institute. (2026). CRISPR clinical trials: A 2026 update. innovativegenomics.org

Queen Mary University of London. (2026, May 30). First results from NHS-Galleri trial presented at international conference. qmul.ac.uk

Industry and Analysis

GRAIL, Inc. (2026, February 19). Landmark NHS-Galleri trial demonstrates a substantial reduction in stage IV cancer diagnoses, increased stage I and II detection of deadly cancers, and four-fold higher cancer detection rate [Press release]. grail.com

Health Affairs Forefront. (2026, June 8). Refining FDA's plausible mechanism framework, part 1: Understanding the draft guidance and clarifying the framework's scope. healthaffairs.org

Sourcing note. Two industry sources are cited above because they are the primary origin of specific dated announcements. Wherever an independent academic or institutional source exists for the same fact, that source has been cited in preference. Claims resting solely on company communications are identified as such in the body text.