Phase 1Cell culture and production (upstream)
01Naked antibody production (CHO fed-batch culture) (production culture, upstream)
The starting step for obtaining the antibody, the foundation of the ADC, consistently at the target yield and quality
- Standard method
- Fed-batch culture in CHO (Chinese hamster ovary) cells
Why it is the standard. CHO is the main host for antibody drugs, and fed-batch culture is the industry standard, reliably delivering titers of several g/L. It is chosen because it can add near-human glycosylation.
Check also. The quality of the foundation antibody (variability in aggregates and glycans) carries straight through to conjugation yield and the homogeneity of the final ADC. Conventional cysteine conjugation needs no additional genetic engineering, but if site-specific conjugation is the goal, the premise changes: an engineered antibody must be prepared at this stage.
About this step and the alternatives
The star of an ADC is the drug, but first a vehicle is needed to carry it. This starting step has factory cells called Chinese hamster ovary (CHO) cells make the antibody protein that serves as the foundation.
It runs as fed-batch culture. As the cells grow they use up their food, so nutrients are topped up during the run, and temperature and feeding are carefully engineered so that antibody is secreted and accumulates in the culture medium.
The key is to get the target amount and quality every time, without drift. Once the antibody's quality is set here, the ceiling on quality for every later step is set too. Precisely because this is the foundation, careful engineering pays off.
The foundation of an ADC is the antibody itself. First, the antibody is made in cells. To obtain it consistently at the target yield and quality, the way the cells are fed and the temperature are carefully engineered.
Alternatives in practice
- Perfusion culture · Sartorius Biostat STR + cell retention device (Repligen XCell ATF 6/10 alternating tangential flow filter), or Cytiva Xcellerex APS — High density, continuous production, shorter residence time. Advantageous for N-1 intensification and continuous production, but operation is complex
- NS0 / Sp2/0 (mouse myeloma) cell expression · Same type of STR / S.U.B. as above (run with media that meet the cholesterol requirement) — Derived from classic hybridomas. Glycosylation differs, so immunogenicity risk assessment is needed
- HEK293 cell expression (transient or stable) · Thermo HyPerforma S.U.B. + Gibco Expi293 system, or wave-type Cytiva WAVE 25 — Human-type glycosylation, rapid expression. For research/early clinical work and difficult-to-express molecules
Harvest and clarification → removal of cells and cell debris by depth filtration
Removes cells and fine particles from the broth to obtain a clarified feed suitable for Protein A loading
- Standard method
- Two-stage depth filtration after centrifugation (coarse → fine) + 0.2 µm bioburden-reduction filtration
Why it is the standard. The de facto standard for clarifying CHO harvest. Cellulose / diatomaceous earth media capture cells, debris and HCDNA at high loading, reducing turbidity before Protein A while maintaining recovery of 95% or more.
About this step and the alternatives
When culture ends, the broth is a cloudy soup packed with CHO cells. It cannot go onto a purification column like this, so this primary recovery (clarification) step first removes the solids to make it clear.
Centrifugation drops the large clumps of cells, followed by two-stage depth filters (coarse → fine) and finally 0.2 µm bioburden-reduction filtration. Step by step, they strain out finer haze and even colloidal impurities.
The result is clear harvest fluid (HCCF). This unglamorous but indispensable preparation step, right before capture, gets the feed ready to load cleanly onto Protein A.
High-density CHO culture broth cannot be loaded onto chromatography as is, so centrifugation and depth filtration remove cells, debris and colloidal impurities to prepare harvested cell culture fluid (HCCF). This primary recovery step is essential right before capture.
Alternatives in practice
- Pretreatment flocculation with Clarisolve + single-stage depth filtration · Merck Clarisolve 60HX + Millistak+ Pod (with a stimulus-responsive flocculant) — Advantageous for high-density perfusion harvests. Pretreatment flocculates particles and reduces filter area
- Primary clarification by continuous centrifugation · Alfa Laval disk-stack centrifuge (CultureOne / Clara 20), or GEA kytero — For large scale and high cell density. Reduces the load on downstream depth filtration
- Acoustic wave cell separation · Pall Cadence Acoustic Separator (BAW standing wave) — For perfusion cell retention and clarification. Low shear, suited to continuous operation
Phase 2Purification (downstream)
03Antibody capture (Protein A affinity purification) (capture step)
Specifically captures the Fc from the large impurity load in the broth, raising purity in one step
- Standard method
- Protein A affinity chromatography (bind and elute)
Why it is the standard. Protein A captures the Fc specifically and achieves high purity and high recovery in 1 step; it is the de facto standard for antibody purification. MabSelect SuRe/PrismA are chosen because their alkali-tolerant ligands stand up well to CIP (cleaning).
About this step and the alternatives
Even the clarified liquid still contains large amounts of impurities besides the antibody, such as host-derived proteins. This capture step picks out only the antibody, as if lifting it out with tweezers.
The key is a resin called Protein A. It binds snugly to only the antibody's tail region (Fc), so when the liquid passes through the column only the antibody adsorbs. After the impurities are washed away, conditions are changed to elute the antibody.
A single operation sends purity soaring: this is the star of downstream processing. Because it is affinity purification that grabs only the intended target, most impurities are left behind here.
The culture broth contains large amounts of impurities besides the antibody. A resin that grabs only the antibody's Fc pulls it out in one go, raising purity.
Alternatives in practice
- Protein L / Protein G capture · Cytiva Capto L / MabSelect VL (κ light chain binding), Tosoh Toyopearl AF-rProtein L-650F, Kaneka KanCapL — For Fc-lacking fragments and κ-chain-containing molecules. For constructs where Protein A does not work
- Cation exchange (CEX) capture · Cytiva Capto S ImpAct / Tosoh Toyopearl GigaCap S-650M + ÄKTA process — Independent of Protein A, low cost. Uses pI differences; requires condition optimization
- Protein A membrane / fiber (convective) capture · Cytiva Fibro PrismA (HiScreen/HiTrap Fibro), Sartorius Sartobind Protein A membrane adsorber — Short residence time, high flow rate; cycling gives higher productivity than resin
Low-pH viral inactivation → inactivation of enveloped viruses by acidic hold of the Protein A elution pool (viral inactivation / VI)
Essential viral clearance step that inactivates enveloped viruses under acidic conditions after Protein A elution
- Standard method
- Adjust to pH 3.3–3.6 with acetic acid and hold at room temperature for 30–60 min or longer, then neutralize with Tris (around pH 5)
Why it is the standard. It takes advantage of the acidic Protein A elution to inactivate retroviruses robustly at pH <3.6, making it the most widespread inactivation method in mAb/ADC processes. It needs no dedicated reagents or equipment and is process-efficient.
Check also. An acidic hold inactivates only enveloped viruses; non-enveloped viruses are removed separately by filtration downstream. Some antibodies undergo acid denaturation or aggregation at low pH, so the hold pH and time are set by measurement at the point that balances viral inactivation and antibody stability.
About this step and the alternatives
The Protein A eluate is acidic to begin with. This safety step turns that property to advantage and knocks out viruses without adding more agents. As a standard precaution, it guards against the unlikely event of contamination.
Acetic acid lowers the pH to 3.3–3.6, and the pool is held as is at room temperature for 30–60 minutes or longer. Retroviruses and other viruses with a membrane (envelope) lose their structure and infectivity in this acidic environment.
Once the treatment has done its job, the pool is neutralized with Tris or similar, back to a pH that does not harm the antibody, and moves on. It forms one side of a double safety net, destroying viruses on a principle different from the removal filtration later on.
Because the Protein A eluate is acidic, it is adjusted to pH 3.3–3.6 with acetic acid or similar and held for a set time to inactivate enveloped viruses (retroviruses, etc.). It is always performed as a standard step to assure viral safety on an orthogonal principle, and after the hold the pool is neutralized with Tris or similar.
Alternatives in practice
- Detergent (solvent/detergent) viral inactivation · Stirred tank + inline addition (Merck Deviron 13-S9 or Triton X-100 alternatives) — For enveloped viruses. For molecules sensitive to low pH. Triton X-100 is being replaced to comply with REACH
- Continuous low-pH viral inactivation · Cytiva ÄKTA process + inline acid addition and tubular reactor (residence-time-controlled coil) — For continuous/intensified processes. No hold tank needed; residence time is tightly controlled
Polishing (removal of impurities, aggregates and viruses)
Removes residual HCP, DNA, aggregates and viruses to finish a safe drug substance antibody
- Standard method
- Ion exchange
Why it is the standard. A 2–3-column train, with AEX flow-through removing DNA/HCP/viruses and CEX/HIC removing aggregates and charge variants, plus viral clearance, is the standard configuration in line with ICH Q5A.
About this step and the alternatives
Capture alone cannot fully remove host cell proteins, DNA, aggregates of antibodies stuck together and trace viruses. This polishing step scrubs them off and buffs the product to a finish.
The tool is ion exchange chromatography. Using differences in molecular charge as the cue, it routes the target antibody and trace impurities separately. Only the intended antibody passes, while unwanted species are held on the resin and separated.
Past this point, the antibody reaches a purity usable as drug substance. Capture does the rough grab, polishing does the finish. This two-stage approach is what finally gets you to a safe drug substance antibody: the climax of purification.
Capture alone leaves host cell proteins, DNA, aggregates and viruses behind. Polishing steps shave off trace impurities to finish a safe drug substance antibody.
Alternatives in practice
- Mixed-mode (multimodal) chromatography · Cytiva Capto adhere (AEX + hydrophobic) / Capto MMC (CEX + hydrophobic), Bio-Rad Nuvia cPrime — Removes aggregates, HCP and leached Protein A in 1 step. High selectivity
- AEX membrane adsorber (flow-through) · Sartorius Sartobind Q / Cytiva Mustang Q, Merck Natrix Q (high-capacity membrane) — Removes negatively charged impurities (DNA/viruses/HCP) in flow-through. Higher flow rate than resin, disposable
- Continuous / multicolumn (MCC) polishing · Sartorius Resolute BioSC Pilot / Cytiva ÄKTA pcc 75 — Improves resin utilization and reduces buffer use. For continuous manufacturing lines
Virus removal filtration → physical removal of non-enveloped and residual viruses by nanofiltration
Essential step that physically removes small viruses such as parvoviruses by size exclusion after polishing
- Standard method
- Constant-pressure nanofiltration with a 20 nm-class hollow-fiber virus removal filter
Why it is the standard. Regenerated cellulose / hydrophilized PVDF hollow fibers remove parvovirus with high LRV while maintaining antibody permeability. The de facto standard for mAb virus removal filtration, with well-established consistency with integrity testing.
With Site-specific conjugation (enzymatic, mTG transglutaminase)
- The mTG method uses native (unreduced) antibody as substrate, so interchain disulfide reduction is not performed. Instead, deglycosylation to expose Q295 is required as pretreatment.
- Method: Trimming (deglycosylation) of the Asn297 N-glycan with EndoS/EndoS2 endoglycosidase. The innermost GlcNAc is left and Asn stays neutral (unlike PNGase F, no deamidation occurs)
- Equipment: Temperature-controlled reactor (thermostatic water bath / jacketed reaction vessel)
- Reagents: EndoS2 (or EndoS) endoglycosidase. TCEP/DTT are not used
- Check also: Methods that use glutamine tags (LLQG or engineered Q-tags) or certain mTG substrates may be able to skip deglycosylation, in which case this step itself is unnecessary
About this step and the alternatives
An acidic hold destroys only viruses that have a membrane. Tough viruses without a membrane resist acid and survive. So this removal filtration step changes the principle and strains them out physically.
The purified antibody is passed at constant pressure through an ultrafine virus removal filter with pores of about 20 nm. The antibody passes through, while larger parvoviruses and the like are stopped by the size barrier.
Inactivation that destroys with acid, and removal that stops by size: layering two moves on different principles assures viral safety. This is the core of viral clearance, finishing the antibody before the drug is attached.
Non-enveloped viruses (parvoviruses, etc.) that low-pH inactivation cannot eliminate are physically removed by size exclusion with a virus removal filter of about 20 nm. It is the core of orthogonal viral clearance and is applied as standard to the purified antibody before conjugation.
Alternatives in practice
- Flat-sheet virus removal filtration (PVDF/PES flat sheet) · Merck Viresolve Pro / Viresolve NFP, Pall Pegasus Prime / SV4 — Alternative to hollow fiber. Simpler equipment setup and easy to combine with prefiltration
- Next-generation hollow-fiber (high-flux) filtration · Asahi Kasei Planova FG1 / S20N (next generation) — Improved throughput and flux. Shorter run time with equivalent parvovirus removal performance
Phase 3Conjugation
07Partial reduction of interchain disulfides (cysteine-based) (partial reduction)
Partially reduces interchain disulfides to expose thiols; precisely controls the starting point for DAR
- Standard method
- Partial reduction of interchain disulfides with TCEP (or DTT)
Why it is the standard. TCEP selectively reduces HH/HL interchain disulfides to generate free thiols and lets the starting point for DAR be controlled, so it is the standard reducing agent for cysteine-conjugated ADCs. The number of thiols (DAR) is engineered through equivalents, time and temperature.
Check also. How far reduction proceeds directly determines the average DAR (drug-to-antibody ratio). If reduction goes too deep, the antibody opens up too much, causing aggregation and chain dissociation, so the equivalents of reducing agent, temperature and time must be engineered within narrow ranges; this step easily becomes the source of lot-to-lot variability. In addition, the standard practice is to prefilter the reducing agent and reaction buffer through 0.2 µm before addition and transfer them aseptically through a closed single-use flow path.
With Site-specific conjugation (enzymatic, mTG transglutaminase)
- Instead of cysteine-maleimide addition, mTG enzymatically catalyzes formation of an isopeptide bond between Q295 and an amine-type linker.
- Method: mTG-catalyzed transamidation (crosslinking of the Q295 side-chain amide and the linker's primary amine). Usually a two-step method: a click handle such as azide/BCN is introduced, then drug carrying DBCO/strained alkyne is attached by SPAAC
- Equipment: pH- and temperature-controlled reactor (near neutral, hold at room temperature to 37 °C)
- Reagents: Microbial transglutaminase (mTG, from S. mobaraensis; activated by removing the pro-sequence with trypsin) + primary amine-containing linker (e.g., azide-PEG-amine, amino-BCN). Maleimide drug-linkers are not used
About this step and the alternatives
Now the drug will be attached to the cleaned-up antibody. This first step creates anchors (reaction sites) on the antibody to hook the drug onto. Think of it as opening just the right number of keyholes for the keys.
Reducing agents such as TCEP or DTT are added to partially open the interchain disulfide bonds that hold the antibody's chains together. This exposes highly reactive thiol groups, which become the attachment points for the drug-linker.
The key is not to open too many. Opening more means more drug attaches, but opening too many makes the antibody fall apart, causing aggregation and quality variability. It is a delicate step that precisely controls the degree of reduction.
As an anchor for attaching the drug, a reducing agent opens the antibody's interchain disulfide bonds to expose thiol groups. Opening too many causes aggregation and quality variability, so the degree of reduction is precisely controlled.
Alternatives in practice
- DTT (dithiothreitol) reduction · Same type of stirred reaction vessel + desalting to remove DTT (Cytiva HiPrep Desalting) — Classic reducing agent. Odor/toxicity, and must be completely removed before the linker reaction
- No reduction needed for lysine (amine) conjugation · (Reduction step skipped) pH-controlled reaction vessel only — Interchain disulfide reduction is not needed when proceeding to NHS ester conjugation
- Site-specific conjugation (engineered cysteine / enzyme tag) · Antibody with engineered cysteines (THIOMAB) + mild reduction/reoxidation reactor — Homogeneous DAR. Requires limited reduction and reoxidation to remove caps
Drug-linker conjugation (conjugation reaction)
Conjugation step that adds drug-linker to exposed thiols and tunes to the target DAR
- Standard method
- Maleimide drug-linker addition (cysteine conjugation)
Why it is the standard. Maleimide-thiol addition is the most widely used conjugation chemistry for cysteine-conjugated ADCs. It has a track record in marketed products such as Adcetris, and it is chosen for its balance of reactivity, selectivity and scalability.
Check also. The drug-linker is poorly water-soluble and is used with organic solvent, so too much solvent causes antibody aggregation and denaturation. Hitting the target DAR is a delicate balance, and even with the average DAR on target, a lot can be a heterogeneous mix of 0, 2, 4 and 8 drugs, so it is controlled together with DAR distribution measurement downstream.
With Site-specific conjugation (enzymatic, mTG transglutaminase)
- No maleimide is used, so there is no maleimide quench with NAC. The reaction is stopped by removing or inactivating mTG.
- Method: Reaction stopped by removing and inactivating mTG (enzyme and excess linker are removed in the following Protein A / chromatography steps). No quench by NAC addition
- Reagents: N-acetyl-L-cysteine (NAC) is not used
About this step and the alternatives
This is the central reaction, finally attaching the drug-linker to the reaction sites (thiol groups) opened in the previous step. Like inserting keys into keyholes, the prepared drug is attached to the antibody.
A part of the drug called the maleimide bonds to the thiol. The drug is poorly water-soluble, so a small amount of organic solvent is used to keep the reaction uniform and attach the drug reliably to the antibody.
The key is neither too much nor too little. Conjugation is tuned to the target drug-to-antibody ratio (DAR = number of drug molecules on 1 antibody). It is the heart of ADC manufacturing, shaping efficacy and safety.
The drug-linker is attached to the exposed reaction sites. The drug is poorly water-soluble, so a small amount of organic solvent is used, and over- and under-conjugation are kept in check to hit the target DAR.
Alternatives in practice
- Lysine-NHS ester conjugation · Stirred reaction vessel controlled at around pH 8 (using NHS-activated linkers such as DM1-SMCC) — T-DM1 type. No reduction needed, but conjugation sites are heterogeneous and the DAR distribution is broad
- Enzymatic site-specific conjugation (transglutaminase/sortase) · Enzyme reaction vessel (microbial transglutaminase mTG, or Sortase A) + temperature control — Homogeneous DAR. Requires introducing a glutamine tag/LPXTG motif
- Re-bridging linkers (disulfide re-bridging) · Stirred reaction vessel after limited reduction + bis-reactive linker (e.g., DBCO/dibromomaleimide type) — Re-bridges interchain disulfides for homogeneous DAR and improved stability
Quench (reaction stop) → inactivation of unreacted maleimide drug-linker / excess payload
Quenches unreacted drug-linker after conjugation, inactivating excess payload before removal
- Standard method
- Maleimide quench by adding N-acetyl-L-cysteine (NAC) (in excess relative to unreacted maleimide)
Why it is the standard. NAC is a water-soluble thiol that rapidly caps unreacted maleimide and forms soluble quench adducts that are easily removed by TFF/chromatography downstream. It is the standard quench reagent in ADC manufacturing.
With Site-specific conjugation (enzymatic, mTG transglutaminase)
- The clearance targets change. The main aim is removal of excess amine linker, mTG enzyme and sugars released by deglycosylation, not free maleimide drug. Protein A recapture is often used as well.
- Method: UF/DF by TFF (+ Protein A affinity repurification as needed to remove excess amino-BCN/mTG)
- Reagents: In the two-step method, clearance also covers excess drug from the click step and copper-free SPAAC reagents
About this step and the alternatives
Even after the drug has been attached, unreacted maleimide and leftover drug-linker remain in the reaction mixture. This quench step inactivates them rather than leaving them be, and stops the reaction cleanly.
A thiol reagent such as N-acetylcysteine (NAC) is added in excess relative to the unreacted material. It neutralizes the remaining reactive groups ahead of time so that no further unwanted conjugation occurs.
If the reaction is not stopped, it causes nonspecific conjugation and aggregation. It is also an unglamorous but effective pretreatment so that the filtration downstream reliably washes out free drug and quench adducts.
After drug-linker conjugation, unreacted maleimide and excess drug-linker are quenched with a thiol reagent such as N-acetylcysteine to stop the reaction. This prevents nonspecific conjugation and aggregation and serves as pretreatment so that free drug and quench adducts are reliably removed by TFF downstream.
Alternatives in practice
- Quench with L-cysteine · Same type of containment stirred reaction vessel — Alternative to NAC. Chosen for removability in downstream UF/DF and pH compatibility
- Capping of residual thiols with N-ethylmaleimide/iodoacetamide · Containment stirred reaction vessel (NEM/IAM addition) — Caps the unreacted free thiols to control DAR/heterogeneity
Phase 6Quality attributes and analytics
12Drug-to-antibody ratio (DAR) and drug load distribution (DAR measurement)
Measures average DAR and load distribution, the most important quality attribute for efficacy and safety
- Standard method
- DAR and drug load distribution by hydrophobic interaction chromatography (HIC)
- Equipment
- HIC-HPLC for DAR
- Reagents and materials
- HPLC mobile phase solvents
Why it is the standard. HIC separates DAR species non-denaturingly by differences in hydrophobicity and can quantify average DAR and distribution, making it the reference method for cysteine-conjugated ADCs. It is an established technique that separates DAR 0/2/4/6/8 with a salt gradient.
Check also. Hydrophobic interaction chromatography (HIC) is the standard method that shows average DAR and drug load distribution at the same time, but with conjugation chemistries whose drug is less hydrophobic, or with site-specific ADCs, the DAR species may not separate fully. In that case, the distribution is cross-checked on a different principle by reversed-phase HPLC or mass spectrometry (intact MS).
About this step and the alternatives
How many drug molecules, on average, are attached per 1 antibody molecule? This most important quality assessment of an ADC measures the drug-to-antibody ratio (DAR) and the spread in drug load (distribution). Too few and it does not work; too many and it tips toward toxicity.
The measurement is hydrophobic interaction chromatography (HIC). It uses the fact that molecules carrying more drug are more hydrophobic, separating molecules by drug load to read both the average and the distribution.
DAR is the first and foremost attribute, shaping efficacy, safety and even behavior in the body. For every lot, it is always confirmed that DAR is within the target range and that molecule-to-molecule variability is acceptable.
The average number of drug molecules attached to 1 antibody molecule (DAR) is the most important quality attribute of an ADC, determining efficacy and safety. Too low and it does not work; too high and it causes toxicity and worse pharmacokinetics.
Alternatives in practice
- DAR by UV/Vis (dual-wavelength) method · UV/Vis spectrometer (Agilent Cary / Thermo NanoDrop) or HPLC-DAD — Calculates average DAR from the absorbance ratio of antibody and payload. Simple, but gives no distribution
- Reversed-phase LC-MS (including native RPLC-MS) · Agilent/Waters UHPLC + C4 reversed-phase column + Q-TOF/Orbitrap MS — Obtains DAR distribution and species identification at once. Effective for hydrophobic payloads
- Native / reduced subunit MS · Waters/Thermo native MS (Q-TOF/Orbitrap) + desalting SEC — Analyzes drug load by subunit (light chain/heavy chain)
Molecular weight and conjugation site identification (intact mass / peptide mapping) (mass spectrometry, peptide mapping)
Structural identification that confirms precisely by mass that drug sits at the right places in the right numbers
- Standard method
- LC-high-resolution MS
- Reagents and materials
- Peptide mapping protease, Analytical antibody enzymes, Disulfide reduction alkylation reagents
Why it is the standard. Peptide mapping LC-MS/MS is the industry standard for conjugation site and occupancy analysis, and intact/subunit MS confirms DAR distribution and molecular weight. Meets structural confirmation under ICH Q6B.
About this step and the alternatives
Is the drug attached at the right places on the antibody, in the right numbers? This step cross-checks precisely, from molecular weight (mass), that the molecule is assembled as intended, matching the actual product against the blueprint.
The tool is LC-high-resolution MS (liquid chromatography + mass spectrometry). Intact mass captures the overall picture of the whole molecule, and peptide mapping cuts it into fragments to pinpoint the conjugation sites.
Looking from both the overall picture and the sites confirms not only how many drugs are attached but also where. It is the precision analysis that grasps the structure accurately and assures the identity of the ADC.
This step confirms precisely, by mass, that the drug is attached at the right places on the antibody in the right numbers. The structure is cross-checked from two sides: the overall mass and the conjugation sites seen after fragmentation.
Alternatives in practice
- Middle-down / middle-up MS · IdeS (Genovis FabRICATOR) digestion + Orbitrap/Q-TOF — Analyzes conjugation sites and DAR at the F(ab')2/Fc subunit level
- N-glycan profiling · Agilent/Waters HILIC-FLR-MS + 2-AB/RapiFluor-MS labeling (Waters GlycoWorks) — Glycoform quantitation. For Fc effector function and CQA control
- Intact MS after deglycosylation · PNGase F treatment + intact measurement on Orbitrap/Q-TOF — Removes glycan heterogeneity and simplifies mass analysis of payload loading
Size variants and aggregates (purity / high-molecular-weight species) (SEC / purity test)
Purity test that measures aggregate levels and monomer fraction, which drive loss of potency and immunogenicity
- Standard method
- Aggregate and monomer quantitation by size-exclusion chromatography (SEC-HPLC)
- Reagents and materials
- HPLC mobile phase solvents
Why it is the standard. SEC is the reference method for purity testing that separates and quantifies aggregates (HMW), monomer and fragments. For ADCs, aggregates <10% is a common guideline, and it meets size heterogeneity assessment under ICH Q6B.
About this step and the alternatives
Conjugation, which attaches the drug, is a reaction that easily causes molecules to stick together (aggregation). This purity assessment step measures how much of those aggregates and size variants there are.
The tool is size-exclusion chromatography (SEC-HPLC). It uses the fact that molecules of different size flow at different speeds, quantifying the fraction of correct monomer and the amounts of oversized high-molecular-weight species and undersized fragments.
Aggregates reduce potency and can also trigger immunogenicity (the risk that the body treats the product as foreign). This purity indicator, directly tied to safety, confirms that monomer is firmly the main species.
Conjugation readily causes aggregation, and aggregates pose risks of reduced potency and immunogenicity. Species are separated by size to measure the monomer fraction and the amounts of high-molecular-weight species and fragments.
Alternatives in practice
- CE-SDS (reduced/non-reduced) · ProteinSimple Maurice S / MauriceFlex (Bio-Techne), SCIEX PA800 Plus — Electrophoretic quantitation of size variants/fragments/purity under SDS
- SEC-MALS · SEC above + Wyatt DAWN/miniDAWN (light scattering) + RI detection — Calculates column-independent absolute molecular weight and the true mass of aggregates
- Analytical ultracentrifugation (AUC-SV) · Beckman Coulter Optima AUC (sedimentation velocity) — Matrix-independent orthogonal confirmation of aggregates. For validating the reference method
Charge variants (charge heterogeneity) (charge heterogeneity / icIEF)
Separates charge changes from deamidation and drug conjugation and measures acidic and basic variant levels
- Standard method
- Charge variant profiling by imaged capillary isoelectric focusing (icIEF)
- Reagents and materials
- Carrier ampholytes pI markers
Why it is the standard. icIEF quantifies acidic/main/basic peaks with high resolution and high throughput (about 10 min/sample) and is the standard for assessing charge heterogeneity of antibodies and ADCs.
Check also. Highly hydrophobic payloads can adsorb to columns and membranes and reduce peak reproducibility. Charge variant assessment mainly targets antibody-derived changes such as deamidation, while the heterogeneity of drug conjugation itself is covered separately by DAR measurement (HIC, etc.); the results need to be read with these roles kept separate.
About this step and the alternatives
This quality assessment measures differences in the charge molecules carry (charge heterogeneity). If charge shifts during manufacturing or storage, behavior in the body and stability can be affected.
Charge changes with chemical changes such as deamidation and with drug conjugation. Imaged capillary isoelectric focusing (icIEF) separates molecules by differences in the pH at which they become electrically neutral (isoelectric point).
It determines how many variants have shifted to the acidic or basic side and confirms that the profile stays constant. This analysis captures hard-to-see chemical changes using charge as the yardstick.
When deamidation or drug conjugation changes a molecule's charge, pharmacokinetics and stability are affected. Molecules are separated by charge differences to determine the amounts of acidic and basic variants.
Alternatives in practice
- Ion exchange chromatography (CEX/AEX-HPLC) · Thermo MAbPac SCX-10 / Tosoh TSKgel CM-STAT + Agilent/Waters HPLC — Separates and quantifies acidic/basic charge variants in a form that can be fractionated
- Conventional cIEF · SCIEX PA800 Plus (cIEF kit) — Classic capillary IEF. When fractionation/collection is needed
- CZE (capillary zone electrophoresis) · SCIEX PA800 Plus / Agilent 7100 CE — High-resolution separation of charge species such as deamidation and C-terminal lysine
Residual free (unconjugated) drug and small-molecule impurities (free drug quantitation)
Measures unconjugated drug and degradants, a cause of systemic toxicity, with high sensitivity to confirm they are within specification
- Standard method
- Quantitation of free drug and related small molecules by reversed-phase HPLC (LC-MS/MS as needed)
- Reagents and materials
- HPLC mobile phase solvents
Why it is the standard. RP-HPLC is the standard method for separating and quantifying free payload and residual linker (generally free drug <2% as a guideline). Trace levels and metabolites are quantified sensitively and specifically by LC-MS/MS.
About this step and the alternatives
This safety assessment measures how much free (unconjugated) drug, left over without attaching to the antibody, remains. Residual free drug causes toxicity by acting throughout the body.
The tools are reversed-phase HPLC and, as needed, LC-MS/MS. They separate and quantify trace free drug and related small molecules such as its degradants with high sensitivity, catching even very small amounts.
Here, whether the filtration in the previous step washed things out properly is confirmed in numbers. It is one of the final checks, directly tied to patient safety, confirming that levels are reliably within specification.
Drug that did not conjugate can cause systemic toxicity if it remains. Trace free drug and its degradants are measured with high sensitivity to confirm they are within specification.
Alternatives in practice
- Solid-phase extraction + LC-MS/MS · Waters Oasis HLB SPE + triple quadrupole LC-MS/MS (SCIEX/Thermo) — Sensitive quantitation by concentrating trace free drug and removing matrix
- Residual solvent GC (headspace) · Agilent 8890 GC + headspace sampler (7697A), FID/MS detection — Quantifies solvents from the conjugation step, such as DMSO/DMAc, in line with ICH Q3C
- ELISA (free/total drug) · Microplate reader (Molecular Devices SpectraMax / Tecan) + anti-payload antibody — Immunoassay of free vs. total payload. For PK/stability assessment
Potency and target binding (potency / binding tests)
Measures target binding and cell-killing ability to assure consistent potency from lot to lot
- Standard method
- Cell-based cytotoxicity (antiproliferation) assay + target binding assay (ELISA / SPR)
- Reagents and materials
- Reporter gene bioassay ADCC CDC, SPR / BLI sensor chips
Why it is the standard. A cytotoxicity assay that reflects the ADC's mechanism of action (target binding → internalization → cell killing) is the core of the potency test for lot release. Combined with binding (ELISA/SPR), it supports the mechanism (ICH Q6B biological activity).
Check also. Cell-based cytotoxicity assays are highly variable, so in practice they are run in parallel with a reference standard and controlled as relative values. For ADCs with a bystander effect, the effect on cells that do not express the target also comes into question, so the assay must be designed around the mechanism of action, together with confirmation of target binding.
About this step and the alternatives
Measuring structure alone cannot tell whether an ADC really works. Ultimately it must be shown by biological activity, that is, whether it can actually kill cancer cells. This potency step confirms that.
Target binding is measured by ELISA or SPR (surface plasmon resonance), and cell-killing ability by a cell-based cytotoxicity (antiproliferation) assay. The ability to catch the target and the ability to deliver toxicity are assessed separately.
Looking at both binding and cytotoxicity assures that the ADC works as intended. This test, which guarantees consistent potency from lot to lot, goes to the very value of the product.
Whether an ADC really works must ultimately be shown by biological activity. Target binding and cell-killing ability are measured to assure that potency is consistent from lot to lot.
Alternatives in practice
- Internalization assay · Fluorescence/confocal imaging (Sartorius Incucyte / pHrodo labeling) or FACS — Assesses ADC uptake and lysosomal trafficking in line with the MOA
- FACS binding assay · Beckman Coulter CytoFLEX / BD FACS flow cytometers — Directly quantifies binding to cell-surface targets. Orthogonal confirmation of SPR/ELISA
- Reporter gene assay · Promega luciferase reporter system + plate reader — Measures receptor binding/signaling MOA-specifically and with high reproducibility
Process-related impurities (HCP / residual DNA / Protein A leaching) (process-related impurity tests)
Sensitively measures residual HCP, DNA and leached Protein A to confirm removal below specification
- Standard method
- Host cell protein (HCP) ELISA + residual DNA qPCR + leached Protein A ELISA
Why it is the standard. HCP ELISA is the standard for ppm quantitation of process-related protein impurities, and residual DNA is quantified sensitively by qPCR. Meets process impurity control under ICH Q6B and regulatory requirements.
About this step and the alternatives
This safety assessment measures not the product itself but the residue of impurities that slipped in during manufacturing. The targets are host cell proteins and DNA from antibody manufacturing and Protein A leached from the resin.
Host cell proteins (HCP) are measured by ELISA, residual DNA by qPCR and leached Protein A by a dedicated ELISA. The best high-sensitivity method is used for each, quantifying down to trace levels.
If these remain, they pose safety risks such as allergic reactions. This test confirms that each step removed them properly and brought them below specification, backing up the reliability of manufacturing.
Host cell proteins and DNA from antibody manufacturing, and Protein A leached from the resin, are safety risks if they remain. Dedicated high-sensitivity assays confirm they have been removed to below specification.
Alternatives in practice
- HCP LC-MS/MS (orthogonal profiling) · Thermo Orbitrap / SCIEX + tryptic digestion, data-independent acquisition (DIA) — Identifies and quantifies individual HCPs that ELISA struggles to capture (lipases, etc.)
- Residual DNA by threshold hybridization · Thermo Threshold System (or an equivalent hybridization method) — Orthogonal confirmation of qPCR. Assesses total DNA independent of sequence
- Control of residual elution/CIP chemicals · Ion chromatography (Thermo Dionex ICS) / LC-MS, conductivity monitoring — Manages clearance of process-derived chemicals such as elution acids, NaOH and detergents
Sterility and endotoxin (general safety) (sterility and endotoxin tests)
Essential step that assures freedom from microbes and pyrogens in an injectable and supports safe administration
- Standard method
- Sterility test + bacterial endotoxins test (LAL / Limulus amebocyte lysate)
Why it is the standard. For bacterial endotoxin, LAL is the reference method in the pharmacopoeias (USP <85>/<161>, Ph. Eur./JP), and the sterility test is a mandatory release test based on pharmacopoeial chapter <71>. Together they assure the general safety of injectables.
About this step and the alternatives
An ADC enters the body directly as an injectable, so there are general safety conditions that can never be skipped. This step provides final assurance that no microbes or pyrogens have contaminated it.
The sterility test checks for microbial contamination. Bacterial endotoxin, which causes fever, is detected by the LAL test using lysate from horseshoe crabs. Even trace amounts are caught.
Only after passing here is it confirmed that the product can be administered safely. No matter how effective an ADC is, it cannot be administered if microbes or pyrogens are present. This is the last gate before the product goes out into the world.
Because the product enters the body as an injectable, freedom from microbes and pyrogens is an absolute requirement. The sterility test and endotoxin test assure that it can be administered safely.
Alternatives in practice
- Methods using recombinant reagents (rFC / rCR) · Lonza PyroGene rFC Assay (fluorescence, Ex/Em 380/440 nm) or FUJIFILM Wako PYROSTAR Neo (recombinant cascade reagent, chromogenic) + absorbance plate reader — Independent of horseshoe crabs, sustainable. Increasingly accepted as an alternative under USP <86>/Ph. Eur.
- Rapid microbiological methods (RMM) · bioMérieux BacT/ALERT、Merck Milliflex Rapid、Rapid Micro Biosystems Growth Direct — Speeds up sterility/bioburden results. For continuous/rapid release
- Monocyte activation test (MAT) · Human monocytes/PBMC + IL-6 ELISA (plate reader) — Detects pyrogens comprehensively (including non-endotoxin pyrogens). Alternative to the rabbit pyrogen test
The reference (“gold standard”) method is the one broadly recognized in published practice and guidelines; your process, product and regulator may justify another. Makers and products are examples, not endorsements. This map is translated from the Japanese edition, where it is updated first (Japanese version).