64 Bispecific / Multispecific Antibody CRO and CDMO vendors

What does it take to develop and manufacture a bispecific or multispecific antibody drug?

A bispecific is not just a harder monoclonal. The moment you ask a single molecule to bind two targets, you inherit a stack of problems that a standard mAb program never sees, and most of them surface early. The first is format. Knobs-into-holes, CrossMab, DART, BiTE, DuoBody, common light chain, scaffold-based designs: the format you pick decides your half-life, your manufacturability, and how ugly the purification gets. A vendor who has only ever made conventional IgG1 will hand you a molecule that looks fine on paper and falls apart at scale.

The engineering and discovery work is where specialists earn their fee. Chain mispairing is the signature failure mode: with two heavy chains and two light chains, the cell can assemble the wrong combinations, and you spend months chasing the correct species out of a soup of half-antibodies, homodimers, and mispaired junk. Good groups design for correct pairing up front (engineered interfaces, charge steering, common light chains) rather than purifying the problem away later. Alongside that sits the developability screen no bispecific should skip: aggregation propensity, thermal and colloidal stability, viscosity at high concentration, and immunogenicity risk from the novel junctions you just created. T-cell engagers add their own layer, since cytokine release and the safety pharmacology around it shape the whole preclinical package.

On the CMC and manufacturing side, the gap between a specialist CDMO and a generalist is widest. Expression is harder because the desired species is often a minority of what the CHO cell secretes, so titers that look healthy can hide a low yield of correctly assembled product. Downstream, you need orthogonal purification (affinity capture plus polishing steps that actually resolve mispaired and aggregated forms) and an analytical package that can prove you have the right molecule: mass spec for chain pairing, intact and reduced mass confirmation, charge variant and size-exclusion profiling, and potency assays that read both arms. A CDMO that has run bispecific cell-line development, scaled the purification, and written the comparability and characterization sections before knows where the bodies are buried. A generalist learns on your program, on your timeline.

How do you choose a CRO or CDMO for a bispecific or multispecific antibody?

Fit to the format matters more than the size of the company. A CDMO with deep AAV experience tells you nothing about whether they can make your CrossMab. Ask for work on your specific format and modality, then push past the marketing slide to the people who would actually run your program. Score two or three vendors against the same written scope using the checklist below, and weight platform fit and analytical depth above headline capacity.

• Relevant platform and track record: real bispecific or multispecific programs in your format family (knobs-into-holes, CrossMab, DART, common light chain, scaffold), not just a generic antibody history. Ask how many reached IND or clinic.
• GxP and analytical capability for this modality: GMP cell-line and process development, plus the analytics that prove chain pairing and purity (LC-MS intact and reduced mass, peptide mapping, charge variants, SEC for aggregates, dual-arm potency). Weak analytics is the most common hidden gap.
• Mispairing and developability control: how they engineer for correct pairing and screen for aggregation, stability, viscosity, and immunogenicity early, rather than discovering a non-developable molecule at the tox-batch stage.
• Capacity and scale: expression titers and correctly-assembled yield for your format, a realistic path from gram-scale tox material to GMP clinical supply, and honest lead times rather than a queue dressed up as availability.
• Regulatory experience: bispecifics drawn IND and IMPD support, comparability and characterization packages, and prior FDA or EMA interactions for this class, including the safety framing for T-cell engagers and cytokine release.
• IP and confidentiality: a CDA before you disclose sequences or format, clear ownership of cell lines and process know-how, and clarity on any platform the vendor licenses in (some bispecific scaffolds carry third-party royalties you inherit).