Evidence atlas · October 29, 2026

THE BODY,
REWRITTEN.

Ten extraordinary advances at the edge of medicine. The results are real; their levels of proof are very different. A visual field guide to what changed, what didn't, and how rare a clinical breakthrough really is.

Explore the ten advances ↘See the historical context

Abstract molecular double helix and linked cells

2026 / RESEARCH BRIEF

10disparate breakthroughs

2FDA decisions featured

824historical oncology trials compared

0claims of universal cures

01 / The signal

A startling season of progress. Not a single kind of proof.

From a twice-as-long median survival in metastatic pancreatic cancer to an in-body gene edit and an engineered pig kidney, the discoveries of 2026 hint at new medical possibilities.

But an approved therapy, an encouraging early trial, and a laboratory mechanism are not equivalent. Each story below carries its evidence label, control-group context, limitations, and links to source material.

02 / The discoveries 10 advances shown

Conceptual oncology illustration; not clinical imaging

01 / ONCOLOGYFDA approved

Twice the median survival. One daily pill.

Daraxonrasib · RASolute 302

13.2 months

Median overall survival; RAS G12 subgroupvs 6.6 months with chemotherapy

In a trial of 500 people with previously treated metastatic pancreatic cancer, the oral RAS inhibitor extended median survival from 6.6 to 13.2 months in the 449-patient RAS G12 subgroup (hazard ratio 0.40). FDA approval followed in August.

Read carefully. Not a cure; results apply to a defined advanced-cancer setting. Severe adverse reactions remain possible.

Approved | randomized Phase 3 · Aug 26, 2026

NEJM randomized trial FDA approval Reporting: The ASCO Post

Conceptual gene editing illustration; not clinical imaging

02 / GENE EDITINGPhase 3 positive

One infusion. Far fewer swelling attacks.

Lonvoguran ziclumeran · HAELO

87% fewer

Monthly hereditary angioedema attacks62% attack- and preventive-therapy-free for six months

A single in-body CRISPR infusion targeting KLKB1 reduced monthly attacks of hereditary angioedema by 87% relative to placebo in an 80-person trial. Over six months, 62% of treated patients had no attacks and used no preventive therapy, versus 11% on placebo.

Read carefully. Durability over many years and rare adverse effects still need observation. The first topline announcement was April 27, two days outside the six-month window.

Randomized Phase 3 · Jun 13, 2026

NEJM randomized trial FDA application status Reporting: Nature Biotechnology

03 / GENE EDITINGPhase 1b

A permanent edit to a cholesterol switch?

VERVE-102 · Heart-2

62% lower

Mean LDL-C at highest tested dose35 people across six dose levels

One infusion of an in-body base editor targeting PCSK9 reduced LDL cholesterol by a mean of 62% at the highest tested dose. In an interim 35-person Phase 1b study, effects appeared durable; 15 participants had at least one year of follow-up.

Read carefully. Not a cardiovascular-outcomes trial. A 62% reduction was the highest-dose group mean, not the result for every participant.

Early human study · May 25, 2026

NEJM early-stage study ClinicalTrials.gov Reporting: STAT

04 / ONCOLOGYPhase 3 positive

A cancer vaccine built for one person.

Intismeran autogene + Keytruda · INTerpath-001

42% lower risk

Recurrence or death; hazard ratio 0.581,137 randomized patients after melanoma surgery

A personalized mRNA neoantigen treatment added to pembrolizumab reduced risk of recurrence or death by 42% relative to pembrolizumab alone after resection of high-risk melanoma. Each vaccine is made using mutations in the recipient’s tumor.

Read carefully. This is a risk reduction, not a 42-percentage-point gain. Overall-survival benefit and regulatory approval are not yet established.

Randomized Phase 3 · Oct 2026

Phase 3 results (Moderna)ClinicalTrials.gov Reporting: The ASCO Post

Conceptual metabolism illustration; not clinical imaging

05 / METABOLISMPhase 3 positive

Weight loss that approaches surgical territory.

Retatrutide · TRIUMPH-1

28.3% lost

Mean body-weight reduction at 80 weeksvs 2.2% on placebo (efficacy estimand)

In the Phase 3 TRIUMPH-1 trial, participants given the highest 12 mg dose had estimated mean weight loss of 28.3% at 80 weeks, compared with 2.2% on placebo. The treatment targets GLP-1, GIP, and glucagon receptors.

Read carefully. The 28.3% result is from the efficacy estimand; an analysis allowing for discontinuation found 25.0%. Gastrointestinal side effects and discontinuation matter.

Randomized Phase 3 topline · May 21, 2026

Lilly Phase 3 results Trial search Reporting: Reuters

Conceptual ai discovery illustration; not clinical imaging

06 / AI DISCOVERYPhase 3 started

AI drug discovery reaches the pivotal stage.

Rentosertib · GENESIS-IPF-3

Phase 3 begun

First patient dosed in pivotal trialIdiopathic pulmonary fibrosis

Rentosertib, discovered with AI-assisted target identification and molecule design, entered a randomized 52-week Phase 3 trial for idiopathic pulmonary fibrosis. The milestone is entry into testing, not an efficacy result.

Read carefully. No Phase 3 success yet. Company descriptions of “first” depend on how AI-designed drug discovery is defined.

Trial initiation only · Sep 10, 2026

First-dose announcement ClinicalTrials.gov Reporting: American Pharmaceutical Review

Conceptual dermatology illustration; not clinical imaging

07 / DERMATOLOGYPhase 3 positive

A new formulation, not a new molecule.

VDPHL01 · Study 304

+36.4 hairs/cm²

Mean increase in non-vellus hair countvs +6.6 hairs/cm² on placebo at six months

An extended-release oral minoxidil formulation met its co-primary endpoints in a randomized 536-man Phase 3 study. Twice-daily treatment increased non-vellus hair count by 36.4/cm², versus 6.6/cm² with placebo.

Read carefully. Second positive registration-directed readout, but Study 302 is designated Phase 2/3. These are company topline data; the medicine is not approved for pattern hair loss.

Randomized Phase 3 topline · Oct 19, 2026

Study 304 results SEC filing Reporting: Fierce Pharma

08 / ONCOLOGYFDA accelerated approval

For some rectal cancers, no surgery required.

Dostarlimab · AZUR-1

71% of 154

No additional treatment and no progressionMSI-H / dMMR stage II–III rectal cancer only

FDA granted accelerated approval to dostarlimab for a specific biomarker-defined form of locally advanced rectal cancer. In AZUR-1, 110 of 154 treated patients (71%) required no other treatment and had no disease progression at data cutoff.

Read carefully. Single-arm evidence and a narrowly eligible population. Close surveillance is essential; longer-term organ preservation and recurrence data are still being collected.

Approved | single-arm Phase 2 · Oct 19, 2026

FDA evidence summary ClinicalTrials.gov Reporting: The ASCO Post

Conceptual transplantation illustration; not clinical imaging

09 / TRANSPLANTATIONCase report

Nine months supported by a pig kidney.

Gene-edited xenotransplantation

271 days

Without dialysis after pig-kidney transplantA single patient; the pig graft eventually failed

Tim Andrews stayed off dialysis for 271 days after receiving a genetically edited pig kidney in January 2025. The graft later failed; he subsequently received a human kidney. The case was reported in The Lancet in September 2026.

Read carefully. This is one patient, not a clinical-trial success rate. It does not demonstrate routine safety, durable function, or generalizability.

One patient; clinical case · Sep 2026

The Lancet case report PubMed abstract Reporting: The BMJ

Conceptual cell biology illustration; not clinical imaging

10 / CELL BIOLOGYLaboratory finding

DNA moved from one cell into another.

Intercellular chromosomal DNA transfer

New mechanism

Chromosomal DNA crossing cell boundariesLaboratory human-cell experiments

Researchers observed micronuclei and chromosome fragments traveling through cytoskeletal nanotubes between adjacent human cells. Some transferred genetic material persisted, became active, and changed recipient-cell traits.

Read carefully. A lab discovery, not a treatment. Its frequency and consequences in intact human tissues remain to be established.

Mechanistic cell biology · May 19, 2026

Cell original research Open-access full text Reporting: Science News

03 / The denominator

How unusual is a successful Phase 3 cancer trial?

A 2026 analysis examined 824 Phase 3 oncology drug trials registered from 2007 to 2023 with available primary-endpoint results. Success meant meeting at least one primary endpoint; it did not mean FDA approval or a particular magnitude of benefit.

Pancreatic cancer15.8%

3 of 19 eligible trials met ≥1 primary endpoint

All oncology43.4%

358 of 824 eligible trials met ≥1 primary endpoint

Melanoma58.1%

18 of 31 eligible trials met ≥1 primary endpoint

04 / Reading the evidence

The rules behind this atlas.

Trial phase is not benefit size.

Phase 3 designates a stage of clinical testing, not guaranteed effectiveness. Positive topline results are less complete than a published, reviewed methods-and-results paper.

Absolute and relative effects differ.

“42% lower risk” is a relative hazard reduction, not an increase of 42 percentage points. Median survival, biomarker changes, and attack counts measure distinct outcomes.

Ten successes are a curated sample.

These were selected because they made headlines. The list does not measure overall medical productivity or prove that a biotechnology “software update cycle” is accelerating.

05 / What it means

The body is becoming more editable. That does not make it software.

Some of the most astonishing results arise from manipulating biological information: changing a gene in the liver, selecting tumor-specific immune targets, or manufacturing an organ from another species. Yet living systems aren't predictable codebases. Their complexity is why safety trials, long-term follow-up, and sober comparisons still matter.

What makes these ten stories extraordinary isn't that uncertainty has vanished. It's how far researchers have moved the frontier while uncertainty remains.

Editorial compilation; current through October 29, 2026. All illustrations are original conceptual schematics, not actual clinical images. Not medical advice. Links open external sources.