00. In Vivo CRISPR Clears Phase 3 (Apr 2026) — Breakthrough

Status: First-ever Phase 3 success for in vivo gene editing; rolling BLA underway | Key source: Intellia HAELO Phase 3

The single biggest development since the last refresh: Intellia's lonvoguran ziclumeran (lonvo-z, NTLA-2002) became the first in vivo CRISPR therapy to succeed in a Phase 3 trial (HAELO, hereditary angioedema, announced Apr 27 2026, published in NEJM). A single 50 mg LNP infusion inactivating the liver KLKB1 gene cut HAE attacks 87% vs placebo (p<0.0001); 62% of treated patients were entirely attack-free and therapy-free vs 11% on placebo; no serious adverse events in the lonvo-z arm across 80 patients. Intellia has begun a rolling BLA (completion 2H 2026) and targets a US launch in 1H 2027.

Why it matters: Casgevy (2023) was ex vivo — cells edited outside the body and re-infused. Lonvo-z proves the harder in vivo path — a systemic infusion that edits the patient's own organ — at registrational scale, and sets the first regulatory precedent for an in vivo CRISPR product. Context: 50+ CRISPR trials are recruiting by mid-2026, but Casgevy remains the only approved CRISPR therapy — lonvo-z would be the second and the first in vivo.

What to watch: FDA's handling of a permanent, irreversible in vivo edit for a serious-but-non-fatal indication. Whether the in vivo LNP-to-liver knockout template generalizes beyond KLKB1. Pricing/reimbursement for a one-time, curative-intent therapy.

---

0. In Vivo Tissue-Specific Gene Editing (Apr 2026) — Breakthrough

Status: Tissue-specific in vivo prime editing demonstrated; ex vivo CRISPR therapies expanding to pediatric | Key sources: RBM20 cardiac prime editing, CASGEVY pediatric, Multiplexed base editing, CRISPR cholesterol trial

The 2026 gene therapy frontier has multiple converging streams:

• In vivo tissue-specific prime editing: RBM20 cardiac prime editing (Feb 2026 bioRxiv) demonstrates dual-AAV delivery of prime editing machinery with editing in heart but no detectable liver editing. Cardiomyopathy phenotype rescue in humanized mouse model.
• Pediatric expansion of approved therapies: Vertex CASGEVY shows first clinical data in children 5-11 with severe sickle cell / beta thalassemia. 1H 2026 global regulatory submissions planned.
• Functional genomics in vivo: Multiplexed in vivo base editing (bioRxiv) pushes base editing toward high-throughput variant-effect screening at organism scale.
• In vivo CRISPR for chronic disease: Verve Therapeutics' Cleveland Clinic cholesterol trial validated single-dose in vivo gene editing for cardiovascular risk.

The general trajectory: ex vivo → in vivo, with tissue specificity as the next major safety milestone. Combined with AI-designed gene editors (OpenCRISPR-1) and DNA-methylation-based gene control (UNSW CRISPR without cutting), the gene-editing platform is becoming a unified toolbox: AI designs → tissue-targeted delivery → precise edit (cut, base-edit, prime-edit, or methylate) → functional validation.

What to watch: Verve cardiovascular Phase 2/3 readouts. Vertex CASGEVY pediatric regulatory submissions through 1H 2026. Beam in vivo programs (sickle cell, AATD). RBM20 cardiac prime editing first-in-human readiness. Whether dual-AAV manufacturing scales economically for tissue-targeted prime editing.

---

Research Frontier

Active Frontiers

1. AI-Designed Gene Editors — Breakthrough

Status: Rapid progress | Key work: OpenCRISPR-1 (Profluent, Nature Jul 2025)

First fully AI-designed CRISPR editor (nuclease + guide RNA + deaminase all generated) successfully edited the human genome. 55.7% on-target, 0.32% off-target (95% reduction vs SpCas9). 403 mutations from SpCas9, 182 from nearest natural variant. Profluent raised $106M in Nov 2025 on the back of this. Publicly released in April 2024.

What to watch: Whether in vivo off-target reduction holds up. Whether the approach extends to base editors, prime editing pegRNAs, zinc fingers, TALENs. Regulatory classification of fully AI-designed biologics. Dual-use biosecurity response.

2. Partial Epigenetic Reprogramming in Humans — Breakthrough

Status: First-in-human Phase 1 starting 2026 | Key work: Life Biosciences FDA approval (Jan 2026), Sinclair 2020 mouse paper, Ledford Nature Apr 2026 critique

First human trial of partial OSK reprogramming (Yamanaka factors minus c-Myc). ~12 glaucoma + up to 6 NAION patients, viral vector delivery to one eye, doxycycline-inducible switch. Readout late 2026 / early 2027.

Opposition voices: Gladyshev (Harvard, "no strong evidence yet"), Williams (eye ≠ systemic), Chandra (Mayo, cell-identity risks). Mechanism appears to be selective demethylation, not embryonic-state rewind.

What to watch: Phase 1 safety readout. Whether OSK is the right cocktail. Whether the approach works outside immune-privileged sites. Doxycycline-switch immunogenicity.

3. Disease-Agnostic Editing — Rapid Progress

Status: Active | Key work: Prime editing suppressor tRNAs (Liu, Nature Nov 2025)

Suppressor tRNAs installed via prime editing read through premature stop codons — addresses ~30% of rare genetic diseases with one strategy. Error rates improved from 1-in-7 to 1-in-101 at normal stop codons. 19 base/prime editing clinical trials underway across 5 countries.

What to watch: Clinical trial results for suppressor tRNA therapies; regulatory pathway for disease-agnostic platform approvals.

4. Epigenetic Editing (No-Cut Gene Activation) — Active

Status: Early stage | Key work: UNSW Sydney / St Jude (Jan 2026)

Demethylation-based gene reactivation without DNA breaks. Applied to Sickle Cell (fetal hemoglobin). Eliminates double-strand break risks. Potentially reversible.

What to watch: Animal model results; durability of demethylation in vivo; head-to-head vs Casgevy.

5. One-Time Cardiovascular Treatments — Active

Status: Phase 1 complete | Key work: Cleveland Clinic ANGPTL3 trial (Nov 2025)

CTX310 LNP-delivered CRISPR: LDL -50%, triglycerides -55% in 15 patients, sustained 60+ days. Targets the leading global cause of death with a one-time treatment.

What to watch: Phase 2/3 design; durability beyond 60 days; pricing and reimbursement for one-time gene therapies.

6. Precision Longevity Framework — Active

Status: Theoretical consolidation | Key work: Baylor Huffington review (Frontiers, Dec 2025)

Synthesizes the field around cell-type-specific interventions. Bulk-tissue aging analysis is obsolete; single-cell atlases identify which populations drive aging in each tissue. Systemic drugs (rapamycin, metformin) have ceilings; targeted delivery enables precision.

What to watch: First human single-cell aging atlases. Functional validation that targeting specific cell populations slows systemic aging. Pharma adoption of cell-type-specific delivery platforms.

7. AI-Accelerated Gene Therapy — Active

Status: Active | Key work: CRISPR-GPT (Stanford, Sep 2025)

LLM copilot trained on 11 years of expert discussions. Novice undergraduate achieved first-attempt gene editing success. Compresses development timelines.

What to watch: Integration with automated wet-lab robotics; biosecurity response; regulatory engagement with AI-designed trials.

8. Base Editing Breaks Through in the Clinic — Breakthrough

Status: De-risked across 3 indications; first genetic correction in patients | Key work: Beam BEAM-302/304/101, Verve VERVE-102 (2026)

Single-base conversion (A→G / C→T) with no double-strand break — "safer CRISPR" — now has human data across AATD (BEAM-302, first-ever genetic correction; durable dose-dependent AAT rise at 15–60 mg), PKU (BEAM-304 FDA IND cleared Jun 2026), sickle cell (BEAM-101, >90% disease-protein reduction sustained ≥36 wk, no busulfan), and cardiovascular (Verve VERVE-102, up to 69% LDL reduction). Almost all via LNP-to-liver in vivo delivery. Distinct from prime editing (search-and-replace) and from Intellia's knockout approach: base editing corrects the disease-causing letter.

What to watch: Multi-year durability of in vivo edits; resolution of the BEAM-101 sickle-cell safety signal (editor vs conditioning); bystander/off-target safety as doses escalate.

9. Long-Read Sequencing Hits the Clinic — Active

Status: Accuracy convergence achieved; single-cell + long-read integrated | Key work: SCLR-seq review (Briefings in Bioinformatics, Dec 2025)

The accuracy tax on long reads is largely paid down — PacBio HiFi at 99.9–99.95% and Oxford Nanopore at up to ~99%. The frontier development is SCLR-seq: single-cell + long-read convergence (MAS-seq / Kinnex) that moves transcriptomics beyond gene-level counts to differential isoform expression, exposing novel isoforms, fusion transcripts, and splicing-derived MHC-I neoepitopes (cancer immunotherapy targets). This is the read-out layer beneath editing (edit validation, variant→function) and longevity (the single-cell atlases the precision-longevity thesis needs).

What to watch: Whether SCLR-seq reaches routine clinical-diagnostic cost/scale; how PacBio and ONT differentiate as accuracy converges; speed of splicing-neoepitope → immunotherapy translation.

Recent Breakthroughs

Knowledge Gaps

Areas where the KB needs more sources — suggested searches for next /kb discover:

• Intellia lonvo-z FDA review + label — The BLA is rolling; the FDA decision and any label/REMS will set the in vivo CRISPR precedent. Suggested: "Intellia lonvo-z FDA approval decision label 2027"
• In vivo base editing durability data — Multi-year follow-up on BEAM-302 (AATD) / VERVE-102 (PCSK9). Suggested: "base editing durability 2-year follow-up in vivo 2026"
• BEAM-101 sickle-cell safety resolution — Was the reported death editor-related or conditioning-related? Suggested: "Beam BEAM-101 sickle cell safety death investigation 2026"
• AI-designed biology beyond CRISPR — AI-designed base editors, prime-editing pegRNAs, or delivery proteins? Suggested: "protein language model de novo design base editor delivery 2026"
• Human single-cell aging atlases — Still the precision-longevity translation bottleneck; now intersects long-read isoform atlases. Suggested: "human aging atlas single cell long-read isoform 2026"
• CRISPR for cancer (in vivo CAR-T) — Beam's anti-CD7 base-edited CAR-T and others are emerging but thin in the KB. Suggested: "in vivo CRISPR base-edited CAR-T clinical trial 2026"
• Non-liver delivery vectors — LNP-to-liver dominates; brain/muscle/lung underrepresented. Suggested: "engineered AAV antibody-conjugated LNP non-liver gene therapy 2026"
• Gene editing pricing and reimbursement — One-time therapies entering the market (lonvo-z) make this concrete, not hypothetical. Suggested: "in vivo gene therapy pricing reimbursement launch 2027"
• Clinical long-read diagnostic validation — Regulatory/CLIA validation of long-read WGS for rare-disease diagnosis. Suggested: "long-read sequencing clinical diagnostic validation rare disease 2026"