Lists like this one usually fail in the same way: they name a mechanism, describe what it might do, and quietly skip the part where a regulator or a randomised trial says whether it worked. So we set a rule before writing. Nothing goes on this list without either a regulatory decision or a named late-stage readout attached to it, with a date. Seven candidates cleared that bar. Several household names did not.
Key takeaways
- Casgevy became the first approved CRISPR medicine in the US on 8 December 2023.
- Lenacapavir for prevention was approved on 18 June 2025; WHO issued guidelines on 14 July 2025.
- Zosurabalpin could be the first new Gram-negative antibiotic class in around 50 years.
- CDER approved 46 novel drugs in 2025, against a historic average of 36 a year since 1993.
The seven, and where each one stands
Read the right-hand column first. It is the difference between a therapy a patient can be prescribed and a therapy a patient can be enrolled in, and that distinction gets lost in almost every article on this subject.
| # | Innovation | Status and date |
|---|---|---|
| 1 | CRISPR cell therapy (Casgevy) | Approved, December 2023 and January 2024 |
| 2 | Bespoke in vivo base editing | First patient treated 2025; FDA pathway February 2026 |
| 3 | Individualised neoantigen therapy | Phase 2b five-year data, ASCO 2026; phase 3 enrolled |
| 4 | CAR-T in autoimmune disease | Early-phase trials; EULAR 2026 readouts |
| 5 | Twice-yearly HIV prevention | Approved 18 June 2025 |
| 6 | Zosurabalpin, new antibiotic class | Phase 3 in about 400 patients |
| 7 | Oral GLP-1 (orforglipron) | Approved 1 April 2026 |
Rewriting the genome, in the clinic and then for one child
1. CRISPR moves from laboratory tool to prescribable medicine
Casgevy, exagamglogene autotemcel, was approved by the FDA on 8 December 2023 for sickle cell disease, making it the first medicine available in the United States that treats a genetic disease using CRISPR gene editing. Approval for transfusion-dependent beta thalassemia followed on 16 January 2024. The therapy is manufactured individually from each patient’s own blood stem cells, edited to raise production of fetal haemoglobin. In 2026 the FDA extended the approval down to children aged two and older.
The significance is categorical rather than incremental. A gene-editing technique described in 2012 became a regulated product in eleven years, which is fast by any historical standard for a new therapeutic modality.
2. A base editor designed for a single patient
In May 2025 a team led by Kiran Musunuru presented, and published in the New England Journal of Medicine, the case of an infant born with severe carbamoyl phosphate synthetase 1 deficiency. An adenosine base editor with a guide sequence written specifically for his mutation was delivered in lipid nanoparticles and dosed when he was seven and eight months old. The therapy went from design to patient in roughly six months.
The regulatory consequence arrived faster than anyone expected. In February 2026 the FDA published detailed guidance for a bespoke approval route, the plausible mechanism pathway, aimed at individualised gene and genome editing therapies for ultra-rare disease. It is a genuine break with the logic of a trial powered on a population, and it is contested for exactly that reason.
Pointing the immune system at a new target
3. Cancer therapy built from one tumour’s mutations
Intismeran autogene, previously mRNA-4157, is an individualised neoantigen therapy: the sequence is derived from mutations found in that specific patient’s tumour. At the 2026 ASCO annual meeting, Moderna and Merck reported five-year phase 2b data in high-risk stage III and IV melanoma following complete resection. Combined with pembrolizumab, it reduced the risk of recurrence or death by 49% and the risk of distant metastasis or death by 59% against pembrolizumab alone, with an encouraging overall survival trend in an exploratory analysis.
A phase 3 adjuvant melanoma study is fully enrolled, and the programme spans nine phase 2 and phase 3 trials across melanoma, non-small cell lung cancer, bladder and renal cancer. Phase 2b results have failed to replicate before, so the phase 3 readout is the one that settles this.
4. CAR-T leaves oncology
Engineered T cells were built to clear cancerous B cells. Turning them on the B cells that drive autoimmune disease is a repurposing that nobody in the field predicted would work as cleanly as it appears to. Data presented at EULAR in 2026 describe next-generation CD19-directed CAR-T and NK cell platforms producing drug-free clinical remissions in highly refractory rheumatic disease, following the case series with follow-up published in the New England Journal of Medicine in 2024.
These are small, early-phase, heavily selected patient groups. The honest description is a signal strong enough to justify large trials, not an established treatment.
Old problems, genuinely new chemistry
5. HIV prevention that requires two appointments a year
On 18 June 2025 the FDA approved lenacapavir, marketed as Yeztugo, for pre-exposure prophylaxis in adults and adolescents weighing at least 35 kg, given as a subcutaneous injection every six months. In the phase 3 PURPOSE 1 and PURPOSE 2 trials, 99.9% or more of participants receiving it remained HIV negative. The WHO issued global guidelines for injectable lenacapavir on 14 July 2025.
Adherence has been the binding constraint on HIV prevention for two decades. Replacing a daily decision with a twice-yearly one addresses the actual failure mode rather than the pharmacology, which is why this belongs on a list about lives rather than about molecules.
6. The first new class against Gram-negative bacteria in about fifty years
Zosurabalpin is a tethered macrocyclic peptide, identified by Roche and Harvard researchers from a screen of close to 45,000 compounds and described in two Nature papers in January 2024. It works by blocking the transport of lipopolysaccharide from the inner to the outer bacterial membrane, a mechanism no existing antibiotic uses, and it is unaffected by known resistance mechanisms in Acinetobacter baumannii. A phase 3 trial comparing it with standard of care in roughly 400 patients with carbapenem-resistant infections is running across Europe, the Americas and Asia.
If it reaches approval it would be the first new treatment for Gram-negative infection in about half a century. Antibiotic candidates fail late and often, so that conditional is doing real work.
7. A GLP-1 medicine you swallow
On 1 April 2026 the FDA approved orforglipron, marketed as Foundayo, a once-daily oral GLP-1 receptor agonist for adults with obesity or overweight with a weight-related condition. It is a non-peptide molecule, which is why it can be taken without the fasting and water restrictions attached to earlier oral formulations. The phase 3 ATTAIN programme showed average weight loss of up to 12.4% at the highest dose over 72 weeks.
The clinical result is modest against the best injectables. The manufacturing and distribution consequence is not: a small-molecule pill scales in a way an injectable peptide does not, and supply has been the limiting factor on this drug class since 2023.
What a list like this systematically leaves out
Approval is not availability. Every therapy above except the pill involves individualised manufacture, cold chain, specialist centres or all three, and the distance between a regulatory decision and a treated patient is measured in years and postcodes. The logistics question is not a footnote to this list, it is the constraint that decides how many lives any of it saves, and it is the same question we looked at through the real impact of blockchain on the pharmaceutical supply chain.
Scale matters too. CDER approved 46 novel drugs in 2025, 34 new molecular entities and 12 biologics, against 50 in 2024 and a historic average of about 36 a year since 1993. Output is healthy. The seven above are exceptional within it, not representative of it.
Our closing position, stated plainly: this decade’s most consequential change is not any single therapy on the list. It is that individualisation, of a gene edit, a cancer vaccine or a cell product, has stopped being a research posture and started acquiring regulatory machinery. Everything else follows from whether that machinery works.
From an approval to a patient
The interesting industrial problem is not discovery. It is how a therapy that exists reaches somebody who needs it.
This article is general information about the pharmaceutical sector and is not medical advice. It does not describe treatments suitable for any individual, and decisions about diagnosis, prevention or treatment belong with a qualified healthcare professional. Sources: FDA and Vertex Pharmaceuticals announcements for Casgevy (exagamglogene autotemcel), approved 8 December 2023 for sickle cell disease and 16 January 2024 for transfusion-dependent beta thalassemia, and the 2026 FDA extension to children aged two and older. Kiran Musunuru and colleagues, New England Journal of Medicine, 2025, and the presentation at the American Society of Gene and Cell Therapy meeting of 15 May 2025, for the patient-specific adenosine base editor in carbamoyl phosphate synthetase 1 deficiency; FDA guidance on the plausible mechanism pathway for bespoke gene editing therapies, February 2026. Moderna and Merck press releases for five-year phase 2b data on intismeran autogene with pembrolizumab presented at the 2026 ASCO annual meeting, giving a 49% reduction in risk of recurrence or death and a 59% reduction in risk of distant metastasis or death. EULAR 2026 reports on CD19-directed CAR-T and NK cell platforms in refractory rheumatic disease, following the CD19 CAR T case series with follow-up published in the New England Journal of Medicine in 2024. Gilead and WHO announcements for lenacapavir (Yeztugo), FDA approval on 18 June 2025 and WHO guidelines on 14 July 2025, with PURPOSE 1 and PURPOSE 2 results. Two Nature papers of January 2024 by Roche and Harvard researchers on zosurabalpin, and CIDRAP reporting on the phase 3 trial in approximately 400 patients with carbapenem-resistant Acinetobacter baumannii. Eli Lilly announcement of FDA approval of orforglipron (Foundayo) on 1 April 2026 and the phase 3 ATTAIN programme showing up to 12.4% average weight loss over 72 weeks. FDA CDER novel drug approvals for 2025: 46 novel drugs, 34 new molecular entities and 12 biologics, against 50 in 2024 and an average of about 36 a year since 1993. Trial results described here are interim or early-phase unless stated as approvals. Updated August 2026.
