Gene Editing Weekly — November 6–13, 2026: risto-cel meets its pivotal endpoint, lonvo-z holds four years, and an AAV cancer signal is confirmed
A pivotal base-editing win in sickle cell disease, four years of an in vivo editor ahead of a March FDA decision, the first confirmed AAV insertional oncogenesis in patients, and a Deaf university's position on who decides what gene therapy is for.
This issue covers November 6–13, 2026, Eastern Time. Beam Therapeutics said its base-edited sickle-cell therapy met the primary endpoint of its pivotal study. Intellia reported four years of follow-up on the in vivo editor now under FDA review. 12
The same week carried the field's sharpest safety reminder in years. The FDA said it had confirmed AAV vector integration into an oncogene in patients treated with Zolgensma, and that the risk extends to both approved onasemnogene products. 3
| Track | What changed | Evidence boundary | Next milestone |
|---|---|---|---|
| Clinical | Risto-cel met freedom from severe vaso-occlusive crises for 12 months, and all primary safety endpoints, at an interim analysis; 49 patients treated, 31 evaluable. 1 | Company topline; the share of evaluable patients reaching the endpoint was not disclosed. 1 | BLA as early as year-end; ASH oral in December. 1 |
| Clinical | Lonvo-z cut attacks 89% across 93 patients through up to four years; 97% never started long-term prophylaxis. 2 | Company-integrated Phase 1/2 and Phase 3 data, not an independent analysis. 2 | PDUFA date March 10, 2027. 4 |
| Clinical | CTX310 cut ANGPTL3 65% and triglycerides 63% at day 30 in 10 patients with severe hypertriglyceridemia. 5 | Open-label, single-arm, 30-day primary window. 5 | Longer follow-up; other Phase 1b cohorts. |
| Regulation | FDA confirmed GLI1 insertional oncogenesis with Zolgensma and extended the neoplasia risk to Itvisma. 3 | Agency safety communication; no enforcement action announced. 3 | Whether labeling and follow-up studies change. |
| Company | Editas cleared Australian ethics and trial notification for Strive, with first EDIT-401 dosing expected this year. 6 | Company statement; no patient has been dosed yet. 6 | Initial safety data in the first quarter of 2027. 6 |
| Research | BRCA1 saturation editing across two cell types rescored thousands of variants and found frequent discordance. 7 | Cell-line assays, not clinical outcomes. 7 | Whether variant databases adopt cross-cell-type scores. |
| Governance | Gallaudet University published an institutional position on gene editing, gene therapy and human diversity. 8 | A university statement; it binds no regulator or sponsor. 8 | How hearing trials define success. |
The window held no gene-editing approval, complete response letter, advisory committee action or BLA acceptance. 9 The only editing-relevant entry on CBER's biologics ledger was the November 6 Zolgensma communication. 9
Clinical
A pivotal endpoint, without the number
Beam's release is unambiguous about the result. BEACON met its primary efficacy endpoint, freedom from severe vaso-occlusive crises for at least 12 consecutive months, and all primary safety endpoints at a prespecified interim analysis. 1
It is silent on the size. As of an October 20 cutoff, 49 patients had been treated, 39 adults and 10 adolescents, and 31 had completed the follow-up the endpoint requires. 1 The release does not say how many of those 31 were crisis-free.
That omission matters because the endpoint is a threshold, not a rate. "Met" says the trial crossed its statistical bar. It does not say whether 60% or 90% of evaluable patients reached 12 months without a severe crisis. 1
The mechanism is base editing at the HBG1 and HBG2 promoters in autologous CD34+ cells. The edit is designed to raise fetal hemoglobin without cutting DNA or silencing BCL11A. 1 Patients still receive busulfan conditioning, so the procedure carries the same transplant burden as other autologous cell therapies.
Beam plans a BLA as early as year-end, which would make risto-cel the first base-editing therapy submitted for approval. 1 Detailed pivotal results are promised for a medical meeting in 2027, with an earlier-cutoff update at ASH in December. 1
The proportion is what to watch. Without it, the result cannot be set against the sickle-cell therapies already approved, and the commercial case rests on the word "met". 1
Four years, and a denominator of 93
Intellia's ACAAI release aggregates 93 patients who received a single 50 mg dose across the Phase 1/2 study and the Phase 3 HAELO trial. 2 Across that set, the attack rate fell 89% from baseline through up to four years, and 97% never started long-term prophylaxis after treatment. 2
Every patient was free of long-term prophylaxis at the latest follow-up. 2 Among the 32 Phase 1/2 patients, at least 90% were attack-free in each month through up to four years of follow-up. 2
Plasma kallikrein, the protein the edit is meant to remove, fell by a mean of 84% at four weeks and stayed there through the follow-up period. 2 That is the mechanistic link the company needs. The DNA edit is producing the protein change, not a downstream coincidence.
The most decision-relevant subgroup is the 35 HAELO patients who entered the trial already on long-term prophylaxis and were randomized to lonvo-z. 2 Their monthly attack rate fell 81% against their own rate on prior prophylaxis. 2
Sixty-three percent reached complete control, more than tripling the attack-free rate they had on prior treatment. 2 Quality of life moved with it: the AE-QoL total score fell 23.5 points from baseline to week 28 against 6.5 points on placebo, where a 6-point change is the accepted threshold for clinical importance. 2
Safety looked stable rather than clean. The most common treatment-emergent events were infusion-related reactions, headache, fatigue and nasopharyngitis, and the company reports no new long-term risks. 2
The regulatory frame arrived the same day. Nature reported that the four-year data keep lonvo-z on course for a March 10 decision. That decision would make it the first CRISPR medicine delivered directly inside the body. 4
The same article recalls the 2025 clinical hold on two other Intellia in vivo studies, after a participant died of liver damage. The FDA lifted the hold this year under added safety measures. 4
The efficacy signal was expected. The clean long-term liver record is the part an in vivo editor still had to show.
Two ways to switch off ANGPTL3
Verve Therapeutics published the preclinical package for its ANGPTL3 base editor in Nature on November 6. 10 The construct packages a guide RNA and mRNA for the ABE8.8 adenine base editor in a GalNAc-modified lipid nanoparticle. 10
The nanoparticle aims at hepatocytes through the asialoglycoprotein receptor. 11
In primary human hepatocytes, peak editing reached 85% to 94%. 11 The authors report no guide-dependent off-target editing at any of about 3,000 candidate sites and no treatment-related structural variants. 10
In non-human primates at 3 mg/kg, liver editing was essentially unchanged between day 49 and the three-year analysis. 11 Circulating ANGPTL3 fell by a time-weighted mean of 94%, triglycerides by 43%, and LDL cholesterol by 13% in normolipidemic animals. 11
The LDL figure is the one to read carefully. In a separate primate model with severely reduced LDL-receptor function, LDL cholesterol fell 46%, which is the setting the program is designed for. 11
Three days later, CRISPR Therapeutics presented Phase 1b data for CTX310. It reaches the same gene with a Cas9 nuclease rather than a base editor. 5
The lipid nanoparticle delivers Cas9 mRNA and a guide RNA, and error-prone repair of the double-strand break disables ANGPTL3. 11
| Verve ANGPTL3 base editor | CRISPR Therapeutics CTX310 | |
|---|---|---|
| Editor | Adenine base editor, ABE8.8 | Cas9 nuclease |
| Delivery | GalNAc lipid nanoparticle | Lipid nanoparticle |
| Evidence | Preclinical: hepatocytes and non-human primates | Phase 1b: 10 patients with severe hypertriglyceridemia |
| Main result | ANGPTL3 −94% (time-weighted mean, NHP); triglycerides −43%; LDL-C −13% in normolipidemic animals, −46% in an LDL-receptor-deficient model 11 | ANGPTL3 −65%; triglycerides −63%; non-HDL-C −51%; ApoB −23%; LDL-C −16% at day 30 5 |
| Durability | Editing stable to three years in NHP 11 | Reductions maintained in seven patients followed at least 90 days 5 |
| Safety | No cumulative liver toxicity at three years; no off-target editing at about 3,000 sites 10 | No related serious adverse events; no clinically relevant transaminase rise; one grade 2 infusion reaction 5 |
Those rows are not comparable cells. One column is an animal package built to show precision and durability. The other is a 30-day human readout in a defined high-risk group. 11
CTX310's cohort was 10 patients with fasting triglycerides above 500 mg/dL. Dosing was fixed at 40 mg below 85 kg or 50 mg at or above it. 5
Median baseline triglycerides were above 1,000 mg/dL. All had diabetes, three had a history of pancreatitis and two had atherosclerotic cardiovascular disease. 5
At day 30, ANGPTL3 fell 65% and triglycerides 63%, with non-HDL cholesterol down 51% and apolipoprotein B down 23%. 5 LDL cholesterol fell 16% from a low mean baseline of 60 mg/dL. 5
No related serious adverse events and no clinically relevant AST or ALT elevations were reported. 5 One grade 2 infusion reaction resolved and the infusion was completed. There were no grade 3 or higher related events and no acute pancreatitis on study. 5
The clinical read-through differs by indication. For severe hypertriglyceridemia, the 63% triglyceride reduction at 30 days is the number that matters. It came without the transaminase signal that would worry a liver-editing program. 5
For familial hypercholesterolemia, the relevant test is LDL lowering in patients with little receptor function. Only the Verve primate model addresses that so far. 11
The off-the-shelf basket
CRISPR Therapeutics also updated its allogeneic CD19 CAR-T, zugo-cel, across 21 patients with refractory rheumatologic disease. 12 In systemic sclerosis, five of six evaluable patients on standard lymphodepletion met the rCRISS-25 response at month three, including two with the stricter rCRISS-50. 12
All nine patients with idiopathic inflammatory myopathy reached a moderate or major Total Improvement Score. Two of three evaluable lupus patients held drug-free DORIS remission past nine months. 12 The company reports no cytokine release syndrome above grade 2, no ICANS or IEC-HS, and no severe infections or graft-versus-host disease. 12
Regulation
The theoretical risk, observed
The FDA's November 6 communication is the most consequential regulatory document of the week. 3 It reports postmarketing malignancies in patients treated with Zolgensma, including confirmed product-related insertional oncogenesis involving GLI1. 3
In two of three GLI1-associated cases, molecular analysis found vector DNA integrated into the host genome, with the construct's promoter inserted upstream of or within the oncogene. 3 Tumors reported include the central and peripheral nervous systems, spinal and paraspinal regions, and kidney. 3
The agency extended the risk to Itvisma, the intrathecal product, because it carries the same vector and construct. No neoplasia cases have been reported after Itvisma to date. 3
The FDA also cited two published reports of AAV integration-driven oncogenesis in other programs, one involving GLI1 in FBX-101 and one involving PLAG1 in RGX-111. 3 It said the benefit-risk balance for both approved products still favors treatment, and that it is evaluating whether regulatory action is appropriate. 3
The mechanism is not new to the label. Both products' prescribing information already lists a theoretical risk of tumorigenicity from vector integration. 3 What changed is that the theoretical case now has confirmed molecular evidence behind it.
That distinction matters for the programs this briefing tracks. Most in vivo editing candidates use lipid nanoparticles that deliver transient mRNA, so the integration risk that applies to AAV gene therapy does not transfer to them directly. 10 Programs that use AAV to deliver editors, particularly those dosing children, inherit the question.
The rest of the ledger
CBER's biologics list gained one editing-relevant entry in the window: the November 6 Zolgensma communication. 9 There was no gene-editing approval letter, complete response letter, advisory committee action or BLA acceptance. 9
The agency's press announcements in the window covered a Factor XIa inhibitor approved November 9, which is not an editing product. 13 A Federal Register search for "gene editing" over the same dates returned no documents. 14
Company and pipeline
Editas, from preclinical to a cleared trial
Editas said on November 9 that it had received human research ethics approval and completed Australia's clinical trial notification for Strive, the Phase 1/2 study of EDIT-401 in heterozygous familial hypercholesterolemia. 6
Dosing is expected this year, with initial safety and tolerability data in the first quarter of 2027 and Part 1 topline in the second half of 2027. 6 That puts the first human readout a few months after the preclinical package it rests on.
That package was presented at AHA. EDIT-401 lowered LDL cholesterol by at least 90% in non-human primates, maintained over 10 months. 6
Roughly 90% mean reductions in Lp(a) and ApoB were maintained over three months. The company describes transient liver enzyme elevations, no histopathologic findings, and no genotoxicity signal in primary human cells. 6
EDIT-401 works by upregulating LDLR rather than disabling a secreted protein, which is a different bet from the ANGPTL3 programs. 6 Its case rests on whether raising receptor expression adds to what existing lipid-lowering drugs already achieve.
The company reported $190.2 million in cash and marketable securities at September 30 and a runway into the second half of 2028. 6 Third-quarter net loss was $23.4 million. 6
A second IND/CTA was accepted for BMS-986527, a CD70-targeted allogeneic CAR-T built with Editas's Cas12a editing, triggering a milestone payment. 6 Bristol Myers Squibb will run the Phase 1 study in clear cell renal cell carcinoma. 6
Research
The same variant, two cell types
Nature Genetics published a saturation genome editing study of BRCA1 on November 11. It scores variants in two human cell models instead of one. 7
The HAP1 arm scored 4,113 variants that had not previously been assayed. The mammary epithelial arm scored 2,157 variants in a non-transformed line. 7
HAP1 scores separated known pathogenic from known benign variants with near-perfect accuracy. 7 The problem appeared when the two maps were overlaid: of 393 variants called loss-of-function in HAP1, only 224, or 57%, were significantly depleted in mammary epithelial cells. 7
Nearly half of the HAP1 loss-of-function variants scored neutral in the second cell type. 7 The authors report that the discordant variants are hypomorphic and confer intermediate cancer risk, a different clinical category from the binary call a single assay suggests. 7
The study also found functional variants outside the usual domains: seven missense variants near splice sites in exon 10, a conserved E2F binding site in the promoter, and a deep intronic variant that creates a pseudoexon. 7
For diagnostics, the result argues against giving a single cell type's functional score decisive weight. 7 For editing programs, it is the same warning one step earlier: an efficiency or specificity result measured in one line may not travel to the cell type being edited.
Ethics and governance
Who decides
Gallaudet University published an institutional position on gene editing, gene therapy and human diversity on November 10. 8 The statement affirms that Deaf lives are an essential part of human diversity and calls for Deaf people to lead decisions about technologies that could change deafness. 8
It is not an opposition statement. Gallaudet acknowledges that gene editing and gene therapy may bring important medical benefits and says its position is not a statement against them. 8 What it rejects is the framing: "Deaf lives are not deficits to be corrected". 8
The timing follows the clinic. The FDA approved Otarmeni, Regeneron's gene therapy for OTOF-related deafness, in April 2026. That condition affects roughly 50 newborns a year in the United States. 15
In August, the first child was treated in a trial for GJB2, the most common genetic cause of deafness in North America and Europe. 8
The operative demand is procedural. Decisions should be made freely, with full information and without coercion, and Deaf scientists, ethicists and families should be at the centre. 8 President Roberta Cordano framed the test as whether progress can be claimed without the people whose lives it seeks to change. 8
For companies, the practical question is what a hearing endpoint measures. A therapy can restore auditory response and still leave a family choosing between a medical outcome and a cultural identity. 15
A governance gap, in a comment
The week's other governance item came from Nature Medicine, not from a bioethics body. 16 A comment by Li Du and colleagues argues that several deaths in Chinese trials of experimental cell and gene therapies appear to reveal gaps in oversight. 16
The authors are at the University of Macau, the Canadian Institute for Genomics and Society and partner institutions. They call for a new layer of independent external review for clinical supervision of high-risk therapies. 16
The comment is behind a paywall, so this briefing has its summary and authorship, not its full argument.
No bioethics body issued a position that this briefing could verify in the window. Both ethics items this week came from academic and institutional statements rather than from a standing body.
Watchlist
- November 14 — ACAAI late-breaking orals present the four-year lonvo-z durability and attack-free data formally. 2
- December 12–15 — ASH: risto-cel's BEACON update takes an oral slot, ahead of the year-end BLA. 1
- Year-end 2026 — Beam's risto-cel BLA submission, which would be the first for a base-editing therapy. 1
- First quarter 2027 — First EDIT-401 safety and tolerability data from Strive. 6
- March 10, 2027 — FDA action date for lonvo-z. 4
Editorial view
The week's items ask the same question in different registers: what does an editing result have to show before it can be acted on? 12
Beam's answer is a threshold met. The trial crossed the endpoint regulators set, and that is a genuine milestone for base editing. But a threshold without a proportion says the bar was cleared, not how far. 1
Intellia's answer is a denominator and a duration. Ninety-three patients, four years, a prior-prophylaxis subgroup and a quality-of-life score with a published threshold let a clinician judge the trade. 2
That is also why the March decision matters beyond one company. A first in vivo approval sets the evidence bar for every liver-directed editor behind it. 4
The Zolgensma communication supplies the counterweight. Permanent expression from an integrating vector bought years of benefit for children with spinal muscular atrophy. It has now produced confirmed insertional oncogenesis. 3
The shift to transient delivery is partly an answer to that risk. The ANGPTL3 programs are the clearest test of whether it works. 510
Gallaudet's statement adds the question the others leave out. Durability is only valuable if the effect is the one patients and families want, and no endpoint can settle that. 15
References
- 1
- 2
- 3
- 4
- 5CRISPR Therapeutics Presents Phase 1b Data for CTX310
ir.crisprtx.com
- 6Editas Medicine Announces Third Quarter 2026 Results and Business Updates
ir.editasmedicine.com
- 7
- 8
- 9What's New for Biologics
fda.gov
- 10
- 11Two Gene Editors Take Different Paths to Silence the Same Lipid Gene
crisprmedicinenews.com
- 12
- 13Press Announcements
fda.gov
- 14Federal Register search results for "gene editing", November 6–13, 2026
federalregister.gov
- 15As gene therapy advances, our values remain clear
gallaudet.edu
- 16
This story was produced automatically by a channel. One sentence is all it takes for Neodrop to keep producing for you.