What does it take to develop and manufacture a Gene Editing (CRISPR-based) drug?
A CRISPR program is really several programs stacked together, and the outsourcing map depends on which delivery route you picked. The biology splits into the editor itself (Cas9, Cas12a, a base editor, or a prime editor), the guide RNA that aims it, and the delivery vehicle that gets both into the right cells. Each of those is a separate body of work with its own suppliers, and a generalist CRO that is strong on small-molecule tox will not have the assays or the regulatory scar tissue for any of it.
Discovery and lead optimization for an editor is mostly guide design and screening: nominating guides against the target locus, measuring editing efficiency, and then doing the work that actually defines a CRISPR program, which is characterizing where else the editor cuts. Off-target analysis is the technical heart of the modality. You are buying GUIDE-seq, CIRCLE-seq or SITE-seq, amplicon deep sequencing of nominated sites, and increasingly an honest look at on-target consequences too: large deletions, chromosomal rearrangements, and translocations that simple indel counting misses. For base and prime editors you add bystander editing and, for base editors specifically, RNA off-target and guide-independent DNA editing. A supplier that only quotes you in-silico off-target prediction is not equipped for this.
Then the work bifurcates by delivery, and this is where CDMO selection gets specific. For in vivo programs the dominant route is now mRNA encoding the editor plus a synthetic or chemically modified guide RNA, co-formulated in a lipid nanoparticle, so you need a CDMO with IVT mRNA capability, capped and modified-nucleotide chemistry, large-scale solid-phase guide RNA synthesis, and LNP formulation and encapsulation. An AAV-delivered editor pulls you into viral-vector manufacturing instead, with all the capsid, full-empty ratio, and titer questions that come with it. Ex vivo programs (the edited cell therapies, including CAR-T made with a knockout) need a cell-therapy CDMO that can run electroporation or LNP delivery into the patient or donor cells under GMP, plus the apheresis-to-product chain. Sitting under all of it is analytical method development for a modality regulators are still writing the rulebook on: editing-efficiency assays, off-target panels qualified to GMP, residual Cas protein and residual nuclease testing, and identity and potency assays that an inspector will accept. Specialist gene-editing CDMOs differ from generalists mainly here, in having built and defended these methods before, and in knowing what the FDA's human gene therapy and genome editing guidances actually expect in a filing.
How do you choose a CRO or CDMO for Gene Editing (CRISPR-based)?
Match the supplier to your delivery route and your edit type first, because the wrong-shaped partner is the most expensive mistake here. An LNP-mRNA in vivo program, an AAV-delivered editor, and an ex vivo edited cell product are three different supply chains with three different supplier shortlists, and a base or prime editor needs characterization assays a plain nuclease shop may not run. Get two or three suppliers quoting against the same written scope, and weigh them on the points below.
Two quieter questions decide more than the headline capabilities. Ask how a supplier handles on-target structural consequences, not just off-target cutting, because large deletions and translocations are where this field has been burned and a thorough partner will raise it before you do. And ask, plainly, how many gene-editing programs they have actually taken to the stage you need, since this modality has far more suppliers claiming capability than suppliers with a real track record.
- Relevant platform and track record: confirm hands-on experience with your specific editor class (nuclease, base editor, prime editor) and delivery route (LNP-mRNA, AAV, ex vivo), and ask for redacted examples of programs they have run to your stage, not just a capability slide.
- Off-target and on-target characterization depth: look for empirical methods (GUIDE-seq, CIRCLE-seq, SITE-seq, deep amplicon sequencing) plus assessment of large deletions, rearrangements, and translocations, and for base editors, bystander and RNA off-target work, never in-silico prediction alone.
- GxP and analytical capability for this modality: GLP for the IND-enabling safety package and GMP for clinical supply, with editing-efficiency, residual nuclease, residual Cas, identity, and potency assays already developed and qualified rather than promised.
- Capacity and scale: realistic queue and lead time, and the right manufacturing fit, IVT mRNA and guide RNA synthesis plus LNP formulation for in vivo, viral-vector suites for AAV, or GMP cell-processing for ex vivo, at the scale your phase needs.
- Regulatory experience: a record of supporting IND or CTA filings for genome-editing products, familiarity with FDA human gene therapy and genome editing guidance and long-term follow-up expectations, and EMA or other regional experience if you are filing there.
- IP and confidentiality: a CDA before you name the target or share guide sequences, clear ownership of program data and any edits made, and a sober read on CRISPR licensing exposure, since the underlying patent landscape can affect freedom to operate for your editor and delivery system.