Three strategies
Add (most common): deliver a working copy of the defective gene. The genome is not edited: a working page is inserted alongside the one that fails. Example: Zolgensma (spinal muscular atrophy).
Silence: switch off a harmful gene. Example: antisense oligonucleotide drugs for muscular atrophy (nusinersen belongs to this family, though it is dosed chronically).
Edit: correct the text in place with CRISPR. Example: Casgevy (sickle cell disease, 2023).
Ledger of approvals
The principal approved medicines (representative, not exhaustive; the definitive source is each agency's website):
| Medicine | Disease | Strategy | First approval |
|---|---|---|---|
| Glybera | Lipoprotein lipase deficiency | Viral vector (AAV) | EU 2012 |
| Kymriah | Acute lymphoblastic leukemia | CAR-T (cells) | FDA 2017 |
| Luxturna | Leber congenital amaurosis | Gene addition (eye) | FDA 2017 |
| Zolgensma | Spinal muscular atrophy | Gene addition | FDA 2019 |
| Hemgenix | Hemophilia B | Gene addition | FDA 2022 |
| Roctavian | Hemophilia A | Gene addition | EU 2022 |
| Casgevy | Sickle cell / beta-thalassemia | CRISPR editing | MHRA 2023 · FDA 2023-24 |
| Lyfgenia | Sickle cell disease | Gene addition | FDA 2023 |
Method note: we leave out dozens of later oncology CAR-Ts to avoid inflating the list; every row can be verified at the cited agency.
Millions per patient
These treatments cost hundreds of thousands to over two million dollars per patient (Zolgensma launched at $2.1M). That is not speculation: it is the cost of manufacturing a medicine for one person in a specialized facility.
European and Latin American health systems negotiate pay-for-performance (refund if it fails) or multi-year installments. Unequal access between rich countries and the rest of the world is, today, the field's biggest ethical problem.