Engineering Guide · Vendor-Neutral

mAb vs Bispecific vs ADC Manufacturing: Format, Cost and Yield Compared

Monoclonal antibody, bispecific antibody, and antibody-drug conjugate manufacturing side-by-side comparison 2H+2L MONOCLONAL CHO fed-batch → Protein A → 2 polish → UFDF mAb 5–8 g/L · $20–150/g 100+ approvals VS KIH · CrossMab BISPECIFIC CHO co-transfect → Protein A → 3 polish (mispair removal) bsAb 1–2 g/L · $100–500/g ~15 approvals VS DAR ≈ 4 ANTIBODY-DRUG CONJUGATE mAb → reduce → conjugate → HIC DAR polish → UFDF ADC $499–778/g · 14 approvals Cytotoxic payload BSL-2/3
Figure 1: The three antibody manufacturing formats. A monoclonal antibody (mAb) is a symmetric IgG built from two identical heavy and light chains — the mature, cheap, high-titer baseline. A bispecific antibody (bsAb) combines two different heavy chains and one or two light chains into an asymmetric heterodimer, using Fc engineering (knobs-into-holes, CrossMab, DuoBody) to force correct pairing. An antibody-drug conjugate (ADC) takes a purified mAb intermediate and chemically links 3–4 (or up to 8) cytotoxic payload molecules per antibody to deliver a targeted cytotoxin.
Quick Verdict

Standard mAb manufacturing is the mature, cheap, high-titer baseline (5–8 g/L CHO fed-batch, $20–150/g COGS, 100+ approvals). Choose a bispecific only when the biology genuinely requires dual epitope engagement — you pay a 2–5× per-gram premium for a mechanism no mAb can deliver. Choose an ADC when you need a targeted cytotoxin: same mAb backbone plus a $500–780/g bioconjugation step dominated by the drug-linker at $200–4,000/g raw material. bsAb yield is limited by chain mispairing; ADC yield is limited by DAR distribution control.

Key differences at a glance

Side-by-side comparison

Factor Monoclonal antibody Bispecific antibody Antibody-drug conjugate
Molecular structure Symmetric IgG (2H + 2L) Asymmetric heterodimer (2H + 1–2L) mAb + covalent drug-linker payload
Typical CHO titer 5–10 g/L (up to 15 g/L) 1–2 g/L (up to 5 g/L optimised) Same as parent mAb (5–8 g/L)
Downstream yield 70–75% 40–60% (mispair removal) 50–65% (mAb) then 60–80% (conjugation)
Downstream complexity Protein A + 2 polish Protein A + 3 polish (HIC/CEX/MMC) mAb train + reduce + conjugate + HIC DAR polish
Cell-line development time 10–12 weeks 12–16 weeks (balanced chain selection) Same as mAb + 3–6 months conjugation PD
Typical COGS per gram $20–150/g $100–500/g (2–5× mAb) $499–778/g (drug-linker dominated)
Facility BSL / containment BSL-1 / OEB 1–2 BSL-1 / OEB 1–2 OEB 4–5 conjugation suite (high-potency)
Key manufacturing risk Aggregation, HCP clearance Chain mispairing, homodimer contamination DAR distribution, free drug, aggregation
Approved products (FDA, 2026) 100+ ~15 14
Modality growth rate Mature (biosimilar pressure) Rapid (immuno-oncology, autoimmune) Rapid (targeted oncology)

Titer and yield ranges reflect published CHO fed-batch platform data. Your vendor's current process economics take precedence. See the biopharmaceutical cost-of-goods guide and the Fermentation Economics Calculator for parameter-level modelling.

Monoclonal antibodies in detail

A monoclonal antibody is a symmetric IgG made from two identical heavy chains and two identical light chains that assemble into a single Y-shaped molecule with a molecular weight of roughly 150 kDa. Modern therapeutic mAbs are produced almost exclusively in Chinese hamster ovary (CHO) suspension cells, engineered for high specific productivity, human-like N-glycosylation, and stable single-copy transgene integration. The FDA has approved more than 100 monoclonal antibody therapeutics (roughly 130+ including biosimilars), spanning oncology (Herceptin, Rituxan, Keytruda, Opdivo), autoimmune (Humira, Remicade, Stelara), infectious disease, cardiovascular, and ophthalmology. The upstream and downstream unit operations are the reference architecture that both bispecific and ADC platforms build on.

How it works

Upstream, a fed-batch culture in a 2,000–20,000 L single-use or stainless steel bioreactor runs 12–16 days at 36.5 °C, with programmed nutrient feeds and a temperature shift to 32–33 °C around day 6 to slow growth and push specific productivity. Modern CHO platforms from Lonza GS Xceed, Thermo Fisher CHO-S, Horizon Discovery (Revvity) CHOZN and CHO-DG44, and Selexis SURE CHO-M deliver 5–10 g/L final titer in a mature stable pool. Downstream is the "platform" mAb train: primary recovery (depth filtration or disc-stack centrifugation), Protein A capture, low-pH viral inactivation, cation-exchange bind-elute or flow-through, anion-exchange flow-through polish, nanofiltration for viral clearance, and ultrafiltration/diafiltration. Overall downstream yield sits at 70–75%. Vendors like Cytiva (MabSelect PrismA), JSR (Amsphere A3), and Tosoh Bioscience (Toyopearl AF-rProtein A) compete on dynamic binding capacity, alkaline stability, and lifetime — decisions that drive per-gram cost of goods.

When mAb wins

Standard mAbs win whenever a single target-engagement mechanism (antagonism, agonism, antibody-dependent cellular cytotoxicity, complement-dependent cytotoxicity, receptor blockade) achieves the intended biology. They dominate on regulatory precedent (well-mapped IND to BLA path), commercial CDMO capacity from Lonza, Samsung Biologics, WuXi Biologics, and Boehringer Ingelheim BioXcellence, biosimilar-driven downward pressure on COGS to $20–50/g at 10,000+ L scale, and the ability to run large-population indications (chronic autoimmune, oncology maintenance) economically. Adalimumab (Humira) at a peak revenue of $20B+ demonstrates the ceiling of the format.

Bispecific antibodies in detail

A bispecific antibody is an engineered molecule that binds two different epitopes with two distinct antigen-binding domains. Approximately 15 bispecific antibodies have received FDA approval as of mid-2026: Amgen Blincyto (blinatumomab), Roche Hemlibra (emicizumab), J&J Rybrevant (amivantamab), Immunocore Kimmtrak (tebentafusp), Roche Vabysmo (faricimab), Lunsumio (mosunetuzumab), Tecvayli (teclistamab), Talvey (talquetamab), Elrexfio (elranatamab), Columvi (glofitamab), Epkinly (epcoritamab), Imdelltra (tarlatamab), Ziihera (zanidatamab), and Bizengri (zenocutuzumab). The dominant clinical applications are T-cell engagers that crosslink CD3 with a tumour antigen (BiTEs, DuoBody, XmAb bispecifics), dual-checkpoint blockade (PD-1 × LAG-3, PD-1 × VEGF), and receptor bridging (Hemlibra bridges Factor IX and Factor X to mimic Factor VIII in hemophilia A).

How it works

The core manufacturing problem is chain assembly. Four different polypeptide chains can be combined in 16 possible ways, only two of which (~12.5%) give the desired asymmetric heterodimer. Format engineering solves this in one of several ways: knobs-into-holes Fc heterodimerization (Roche/Genentech, invented in the mid-1990s) engineers a "knob" mutation in one heavy chain CH3 and a complementary "hole" in the other; Roche CrossMab swaps the CH1 and CL domains in one Fab arm to prevent light-chain mispairing; Genmab DuoBody uses controlled Fab-arm exchange between two separately expressed IgGs; Xencor XmAb combines charge-pair mutations with knobs-into-holes; and common-light-chain platforms select for a single shared light chain that pairs correctly with both heavy chains. CDMO platform technologies include WuXi Biologics WuXiBody and Lonza bYlok. Titer typically lands at 0.6–2.2 g/L in stable CHO pools; Gong and Wu (2023, Antibody Therapeutics) demonstrated ~5 g/L for a Fabs-in-tandem IgG by optimising the transfection ratio of the two heavy chains. Downstream needs an extra 1–2 chromatography steps beyond the standard mAb polish to strip homodimers, mispaired species, and half-molecules — differential Protein A elution, HIC, mixed-mode resins, or cation-exchange in bind-and-elute mode.

When bispecific wins

Bispecific antibodies win when the biology genuinely requires simultaneously engaging two epitopes. T-cell engagers are the largest clinical category: teclistamab (Tecvayli, BCMA × CD3) and elranatamab (Elrexfio, BCMA × CD3) in multiple myeloma, epcoritamab (Epkinly, CD20 × CD3) and glofitamab (Columvi, CD20 × CD3) in DLBCL, tarlatamab (Imdelltra, DLL3 × CD3) in small-cell lung cancer, and mosunetuzumab (Lunsumio, CD20 × CD3) in follicular lymphoma. Non-T-cell-engager wins include Vabysmo (VEGF-A × Ang-2) in wet AMD, Rybrevant (EGFR × MET) in EGFR-mutant NSCLC, and Hemlibra (FIX × FX) in hemophilia A. Bispecifics accept a 2–5× per-gram cost premium over standard mAb because the format is what unlocks the mechanism.

Antibody-drug conjugates in detail

An antibody-drug conjugate is a mAb (or occasionally a bispecific) covalently linked to a cytotoxic payload via a chemical linker. The FDA has approved 14 ADCs as of mid-2026, dominated by four toxin classes: auristatins (MMAE, MMAF used in Adcetris, Padcev, Polivy, Tivdak, Blenrep, Emrelis), maytansinoids (DM1 in Kadcyla, DM4 in Elahere), calicheamicins (Mylotarg, Besponsa), pyrrolobenzodiazepine (PBD, in Zynlonta), and topoisomerase-I inhibitors (deruxtecan/DXd in Enhertu and Datroway, SN-38 in Trodelvy). Enhertu (trastuzumab deruxtecan) leads the category at ~$3.75B in 2024 revenue, followed by Adcetris (~$1.9B), Padcev (~$1.6B), and Trodelvy (~$1.3B).

How it works

ADC manufacturing has two independent halves. First, the naked mAb intermediate is produced with the same CHO fed-batch and Protein A downstream train as any therapeutic monoclonal — typical CDMO price for a GMP mAb intermediate suitable for conjugation is ~$300/g. Second, the mAb undergoes bioconjugation in a dedicated conjugation suite (OEB 4–5 containment, high-potency): for interchain cysteine conjugation (Adcetris, Kadcyla, Padcev) the four interchain disulfides are partially reduced with TCEP to expose 4 or 8 free thiols per mAb, then reacted with a maleimide-capped drug-linker in aqueous buffer at 4–25 °C; for site-specific engineered platforms (Enhertu, Trodelvy, Datroway) the antibody has engineered cysteines or unnatural amino acids at defined positions that permit higher-purity DAR 8 species. The reaction is quenched, excess drug-linker is removed by tangential-flow filtration or size-exclusion, and DAR distribution is polished by hydrophobic interaction chromatography (HIC) or ceramic hydroxyapatite. DAR characterisation uses reversed-phase HPLC and native mass spectrometry. Independent cost-of-goods models (Biopharm Services, 2018) put the total ADC COGS at $499.70/g in an integrated mAb + conjugation facility and $777.50/g in a dedicated conjugation-only facility, with drug-linker raw material at $200–$2,000/g the dominant cost driver.

When ADC wins

ADCs win when the tumour target is well-validated for internalisation (HER2, TROP2, Nectin-4, folate receptor alpha, BCMA, CD22, CD33, CD30) and the naked mAb has no direct-kill mechanism against the tumour phenotype. The therapeutic thesis is that a targeted cytotoxin delivers a lethal payload dose selectively to antigen-positive cells while sparing systemic exposure. The commercial market rewards this — Enhertu alone has expanded its label across HER2-positive breast, gastric, and lung cancers, and HER2-low breast cancer. CDMO capacity from Lonza Bioconjugates, WuXi Biologics ADC, Abzena, Piramal Pharma Solutions, and Sartorius (formerly Polyplus) is expanding fast, though the OEB 4–5 containment requirement means dedicated conjugation suites are capital-intensive and lead times can stretch to 12–18 months for a new facility slot.

Pros and cons

Monoclonal antibody

Advantages

  • Mature platform: 100+ FDA approvals, well-mapped IND-to-BLA path, extensive biosimilar precedent.
  • Highest titer and lowest per-gram COGS: 5–10 g/L fed-batch, $20–150/g at commercial scale.
  • Plug-in downstream: platform Protein A + 2 polish + UFDF applies to almost any IgG.
  • Largest CDMO capacity pool: Lonza, Samsung Biologics, WuXi Biologics, Boehringer Ingelheim, Catalent, Fujifilm Diosynth are all bookable in months.

Disadvantages

  • Single mechanism of action — cannot bridge two epitopes or deliver a cytotoxin.
  • Biosimilar erosion after loss-of-exclusivity compresses revenue (Humira, Rituxan, Herceptin).
  • Direct-kill activity is weak against solid tumours without payload delivery.
  • High-dose subcutaneous formats need high-concentration formulation development to hit under 2 mL injection volumes.

Bispecific antibody

Advantages

  • Unlocks mechanisms no mAb can deliver: T-cell engagement, dual-checkpoint blockade, receptor bridging.
  • Off-the-shelf inventory: unlike CAR-T, no patient-specific manufacturing wait.
  • Rapid growth category: ~15 approvals in the last 10 years, immuno-oncology and autoimmune pipelines expanding.
  • Reusable IgG-like production infrastructure (fill-finish, cold chain).

Disadvantages

  • Chain mispairing caps stable-pool titer at 1–2 g/L versus 5–8 g/L for mAb; 2–5× per-gram COGS.
  • Extra downstream polishing steps to strip homodimers, mispaired species, and half-molecules.
  • Cytokine release syndrome risk for T-cell engagers requires step-up dosing schemes and REMS.
  • Format IP is fragmented across knobs-into-holes, CrossMab, DuoBody, XmAb — licensing negotiations add cycle time.

Antibody-drug conjugate

Advantages

  • Delivers a cytotoxic dose selectively to antigen-positive cells: Enhertu label expansions demonstrate the commercial ceiling.
  • Reuses the standard mAb upstream and Protein A capture — the conjugation step bolts on downstream.
  • Fastest-growing modality in oncology by product count and revenue.
  • Site-specific conjugation platforms (engineered cysteine, unnatural amino acid, glycan-engineering) enable narrow DAR 8 species with up to 80% homogeneity.

Disadvantages

  • OEB 4–5 conjugation suite is capex-intensive; CDMO capacity constrained with 12–18 month lead times.
  • Drug-linker raw material dominates COGS at $200–$4,000/g; total ADC COGS ~$500–780/g.
  • DAR distribution and free-drug residuals require dedicated in-process analytics (RP-HPLC, native MS, LC-MS free drug).
  • Payload safety and containment discipline required across manufacturing, sampling, and shipping.

Which format should you choose?

Pick based on the dominant constraint: mechanism required, target-population size, and cost of goods ceiling.

Large-population chronic indication

Autoimmune (Humira class), oncology maintenance, ophthalmology, cardiovascular. Per-gram COGS dominates programme economics at 100+ mg/kg cumulative dosing.

Choose mAb

Mechanism needs two epitopes at once

T-cell engagement (BCMA × CD3, CD20 × CD3), dual checkpoint (PD-1 × LAG-3, PD-1 × VEGF), receptor bridging (Hemlibra). No mAb can deliver this biology.

Choose Bispecific

Targeted cytotoxin for solid tumour

Well-validated internalising target (HER2, TROP2, Nectin-4, FRα), naked mAb has weak direct kill, dose window supports $500–780/g COGS at 5–20 mg/kg.

Choose ADC

First programme — no established platform

Cheapest capex, largest CDMO capacity, deepest regulatory precedent. Start with mAb unless the mechanism specifically forces you off the format.

Choose mAb

Real-world use cases

Approved products illustrate why the industry has converged on each format for specific problems.

mAb · Anti-TNFα · Autoimmune
Humira (adalimumab) at 10,000+ L scale

AbbVie's Humira peaked at $20B+ annual revenue on the strength of the mAb platform's per-gram economics. Now under biosimilar pressure from 10+ approved biosimilars; the format itself is what enabled the scale-out.

Bispecific · CD3 × BCMA · Myeloma
Tecvayli (teclistamab) — J&J

The first BCMA × CD3 T-cell engager approved for multiple myeloma. Off-the-shelf alternative to autologous BCMA CAR-T (Carvykti) at a fraction of the vein-to-vein time and lower per-episode cost.

Bispecific · FIX × FX · Hemophilia A
Hemlibra (emicizumab) — Roche

A non-oncology bispecific: bridges Factor IX and Factor X to substitute for missing Factor VIII in hemophilia A. Subcutaneous, monthly dosing. The classic example of a mechanism no mAb can deliver.

ADC · HER2 · Solid tumours
Enhertu (trastuzumab deruxtecan) — Daiichi/AZ

DAR 8 deruxtecan payload with a cleavable linker. Expanded from HER2+ breast into HER2-low, gastric, NSCLC. ~$3.75B 2024 revenue and the reference commercial-scale site-specific ADC.

Modelling per-gram COGS across formats?

The Fermentation Economics Calculator estimates cost of goods for CHO fed-batch mAb and bispecific runs by titer, downstream yield, and batch count. Combine with the biopharmaceutical COGS guide to sanity-check per-gram economics before locking a platform decision.

Open the Fermentation Economics Calculator

Cost and lifecycle considerations

Cost drivers are structurally different for each format

mAb COGS is dominated by fixed facility overhead and consumables amortised across large fed-batch runs. Bispecific COGS is dominated by low titer and extra downstream yield loss from mispair removal. ADC COGS is dominated by drug-linker raw material at $200–$4,000/g. Optimising the wrong lever wastes engineering effort.

Standard mAb manufacturing at commercial scale (10,000–15,000 L bioreactors, 5–8 g/L fed-batch, mature downstream) sits at $20–150 per gram cost of goods sold. Continuous mAb bioprocessing benchmarks around $51/g against $99/g for fed-batch in published economic analyses. The Bill & Melinda Gates Foundation and LifeArc have publicly targeted a $10/g platform for global-health mAbs — the theoretical floor for the format.

Bispecific antibodies typically run 2–5× the per-gram COGS of a standard mAb. The dominant drivers are lower titer (1–2 g/L versus 5–8 g/L, halving the mass produced per batch), extra downstream polishing to remove mispaired species and homodimers (each additional column adds yield loss and buffer consumption), and smaller batch sizes reflecting the earlier commercial stage of the modality. As bispecific titers improve toward 5 g/L in optimised platforms, per-gram economics will compress toward the mAb baseline.

ADC total cost of goods sits at $499.70/g in an integrated mAb + conjugation facility and $777.50/g in a dedicated conjugation-only facility that buys the mAb intermediate at ~$300/g (Biopharm Services, 2018 cost model). The drug-linker raw material is the dominant cost driver at $200–$4,000/g depending on toxin class — auristatins (MMAE, MMAF) at the lower end, deruxtecan and pyrrolobenzodiazepines at the higher end. Yield in the conjugation step itself is typically 60–80%, and DAR polishing on HIC adds another 5–15% yield loss.

Cost component Monoclonal antibody Bispecific antibody Antibody-drug conjugate
CHO fed-batch upstream~$5–15/g~$20–60/g~$5–15/g (mAb intermediate)
Protein A + polish downstream~$10–30/g~$40–120/g~$10–30/g (mAb intermediate)
Drug-linker raw materialN/AN/A~$200–2,000/g of ADC
Conjugation + DAR polishN/AN/A~$100–200/g of ADC
QC and release testing~$5–20/g~$15–50/g~$20–60/g
Facility overhead per gram~$5–50/g~$20–200/g~$50–200/g (OEB 4–5)
Total COGS per gram~$20–150/g~$100–500/g~$499–778/g

Vendor landscape

Major commercial CHO platforms, CDMOs, and bioconjugation specialists in each format.

Monoclonal antibody platforms and CDMOs

Bispecific antibody platforms and CDMOs

ADC bioconjugation CDMOs

Frequently asked questions

What is the difference between mAb, bispecific antibody, and ADC manufacturing?
All three start from a CHO cell culture upstream, but they diverge sharply in complexity. Monoclonal antibody (mAb) manufacturing produces a single symmetric IgG from two identical heavy and two identical light chains; the process is CHO fed-batch to 5–8 g/L, Protein A capture, two polishing steps, viral clearance, and UFDF. Bispecific antibody (bsAb) manufacturing produces an asymmetric IgG-like molecule from two different heavy chains and one or two light chains — only ~12.5% of the possible chain combinations give the correct product without design intervention, so titers land at 1–2 g/L (up to 5 g/L in optimised lines) and downstream needs extra polishing to remove homodimers, mispaired species, and half-molecules. ADC manufacturing takes a purified mAb (or occasionally a bsAb) as intermediate and adds a bioconjugation step: reduce the interchain disulfides, react with a drug-linker at a target drug-to-antibody ratio (DAR) of 3–4, purify by HIC or SEC to a narrow DAR distribution, then formulate. ADC cost per gram is dominated by the drug-linker at $200–$4,000/g raw material.
How does bispecific antibody titer compare to standard mAb titer in CHO?
Standard monoclonal antibody manufacturing in CHO fed-batch routinely achieves 5–10 g/L at commercial scale, with 10–15 g/L achievable in optimised platform processes. Bispecific antibodies typically land at 0.6–2.2 g/L in stable CHO pools, with optimised stable lines reaching ~5 g/L (published example: Gong and Wu 2023, Antibody Therapeutics, showed a Fabs-in-tandem IgG produced at ~5 g/L with 97% purity after transfection strategy optimisation). The gap comes from the fundamental chain-pairing problem: four polypeptide chains can combine 16 ways, of which only two (~12.5%) give the desired heterodimer, and cell-line development must select clones that express chains in the right ratio. Rescuing yield requires design innovations like knobs-into-holes Fc engineering, common-light-chain platforms, or single-chain fusion formats.
How much does mAb manufacturing cost per gram compared to bispecific and ADC?
Standard mAb manufacturing at commercial scale (10,000–15,000 L bioreactors, 5–8 g/L fed-batch, mature downstream) sits at $20–150 per gram COGS. Continuous mAb bioprocessing benchmarks around $51 per gram versus $99 per gram for fed-batch. Bispecific antibodies typically run 2–5× the per-gram COGS of a standard mAb, driven by lower titer (1–2 g/L versus 5–8 g/L), extra downstream polishing to remove mispaired species and homodimers, and smaller batch sizes reflecting the earlier commercial stage of the modality. ADC manufacturing cost per gram of final drug substance sits at roughly $500–780/g in independent cost-of-goods models (Biopharm Services, 2018): $499/g in an integrated facility that runs mAb production and conjugation on-site, and $778/g in a dedicated conjugation-only facility that buys the mAb intermediate at ~$300/g. Drug-linker raw material dominates ADC COGS at $200–$4,000/g depending on the toxin class (auristatins, maytansinoids, deruxtecans, calicheamicins).
How many mAb, bispecific, and ADC products are FDA approved?
As of July 2026 the FDA has approved more than 100 monoclonal antibodies (roughly 130+ across all mAb sub-classes including biosimilars), approximately 15 bispecific antibodies (Blincyto/blinatumomab, Hemlibra/emicizumab, Rybrevant/amivantamab, Kimmtrak/tebentafusp, Vabysmo/faricimab, Lunsumio/mosunetuzumab, Tecvayli/teclistamab, Talvey/talquetamab, Elrexfio/elranatamab, Columvi/glofitamab, Epkinly/epcoritamab, Imdelltra/tarlatamab, Ziihera/zanidatamab, Bizengri/zenocutuzumab, and others), and 14 antibody-drug conjugates (Mylotarg/gemtuzumab ozogamicin, Adcetris/brentuximab vedotin, Kadcyla/ado-trastuzumab emtansine, Besponsa/inotuzumab ozogamicin, Polivy/polatuzumab vedotin, Padcev/enfortumab vedotin, Enhertu/trastuzumab deruxtecan, Trodelvy/sacituzumab govitecan, Blenrep/belantamab mafodotin, Zynlonta/loncastuximab tesirine, Tivdak/tisotumab vedotin, Elahere/mirvetuximab soravtansine, Datroway/datopotamab deruxtecan, Emrelis/telisotuzumab vedotin). mAbs are the mature, dominant modality; bispecifics and ADCs are the growth categories.
Why is bispecific antibody purification more complex than mAb purification?
A monoclonal antibody purifies cleanly on Protein A because Protein A binds the Fc region of any IgG with high affinity, and the only significant impurities to remove are CHO host-cell proteins, DNA, aggregates, and leached ligand. A bispecific antibody run comes off the bioreactor with a mixture of the desired heterodimer plus multiple product-related impurities: two homodimers (each parental chain paired with itself), mispaired light chains, half-molecules missing one heavy chain, and aggregates. Because these impurities share the Fc region and much of the surface chemistry with the product, they cannot be removed on Protein A alone. Purification typically needs an extra 1–2 chromatography steps beyond the standard mAb polish: differential Protein A elution to strip mispaired species that bind more weakly, hydrophobic interaction chromatography (HIC), mixed-mode resins, or cation-exchange in bind-and-elute mode. Overall downstream yield drops from 70–75% for a mAb to 40–60% for a bispecific.
What is the drug-to-antibody ratio (DAR) in ADC manufacturing and why does it matter?
The drug-to-antibody ratio (DAR) is the average number of cytotoxic payload molecules conjugated to each antibody. Commercial ADCs target a DAR of 3–4 for interchain cysteine conjugation (Adcetris, Kadcyla, Padcev) or a fixed DAR of 8 for site-specific engineered cysteine platforms (Enhertu, Trodelvy, Datroway with the deruxtecan payload). DAR matters for three reasons: potency scales roughly linearly with payload count, so under-conjugated ADC (DAR 0–2) is subtherapeutic; pharmacokinetic clearance accelerates sharply at DAR 6+, so over-conjugated species clear faster from circulation and lose exposure at the tumour; and aggregation risk increases with DAR because each conjugated site increases surface hydrophobicity. Conjugation processes must therefore hit a narrow target DAR distribution (typically DAR 3.5–4.2 for interchain-cysteine platforms) with less than 5–10% of the material outside the DAR 2–6 window. Modern site-specific platforms (Enhertu's Cys-based site-specific conjugation, engineered cysteine, unnatural amino acid, glycan-engineering) can hit up to 80% single DAR species.
Can the same CHO cell line make a mAb, a bispecific, and an ADC?
The upstream CHO host cell line (CHO-K1, CHO-DG44, CHO-S, Lonza GS-CHOK1SV) can be used for all three modalities — the difference is in the transgene, the vector design, and the cell-line development strategy. A standard mAb needs a single vector encoding one heavy and one light chain, and cell-line development selects clones with high specific productivity (qP) and stable single-copy integration. A bispecific antibody typically needs two heavy chains and one or two light chains, either co-transfected on separate plasmids or on a single polycistronic vector with 2A peptides or IRES elements; cell-line development additionally selects for balanced chain ratios (usually enriched for one chain 1.5–2× to force correct pairing). An ADC uses the same CHO clone as the parent mAb; only the downstream adds the bioconjugation step. So the CHO platform is shared, but bispecific cell-line development takes 2–4 weeks longer than mAb (12–16 weeks vs 10–12 weeks) and ADC adds 3–6 months of conjugation process development on top of the mAb timeline.
When should I choose bispecific over mAb or ADC for a new programme?
Choose a standard mAb when a single target-engagement mechanism (antagonism, agonism, ADCC, CDC) achieves the biology you need, regulatory precedent matters (over 100 approvals), and per-gram cost minimisation is critical for a large-population indication (chronic autoimmune, oncology maintenance). Choose a bispecific when the biology genuinely requires simultaneously engaging two epitopes on one or two targets — T-cell engagers (BiTEs, DuoBody, XmAb bispecifics) crosslinking CD3 with a tumour antigen, dual-checkpoint blockade (PD-1 × LAG-3, PD-1 × VEGF), or receptor bridging (Hemlibra bridges Factor IX and Factor X to mimic Factor VIII in hemophilia A). Accept the yield and cost penalty (2–5× per gram) as the price of a mechanism you cannot get any other way. Choose an ADC when the tumour target is well-validated for internalisation (HER2, TROP2, Nectin-4, folate receptor alpha, BCMA, CD22, CD33), the payload class has demonstrated activity in that tumour type, and the therapeutic window supports a $500–780/g COGS at typical 5–20 mg/kg dosing. ADCs are the fastest-growing modality in oncology because they deliver a cytotoxic dose selectively where a naked mAb has no direct-kill mechanism.

Resources and references