ADC Formulation in 2026: Five Critical Challenges Every Developer Must Solve

SEO Keywords: ADC formulation challenges 2026, antibody drug conjugate formulation, drug antibody ratio DAR control, ADC aggregation stability, ADC linker chemistry formulation, HPAPI containment ADC manufacturing, ADC cold chain storage, site-specific conjugation ADC, ADC subcutaneous formulation, ADC CDMO manufacturing challenges, ADC clinical pipeline 2026

There are 614 ADC drug candidates currently in clinical development. Twenty-eight have been approved by regulatory agencies globally. The most recent, AbbVie's Emrelis for non-squamous non-small cell lung cancer, was approved by the FDA in June 2025.

The pipeline has never been larger. The science has never been more ambitious.And the formulation problems have never been harder to hide.

ADC formulation is the discipline that determines whether a molecule withexceptional target selectivity actually survives manufacturing, storage,transit, and infusion with its structure, its payload, and its safety profileintact. When formulation goes wrong, it does not usually announce itself with afailed batch. It announces itself with unexpected clinical toxicity, ashelf-life that does not support commercial supply, an aggregation profile thatregulatory reviewers flag as inadequately characterised, or a scale-up transferthat behaves nothing like the process that produced the clinical data.

Here are the five formulation challenges that are costing ADC programmes time,money, and in some cases, the programme itself.

Challenge 1: The DAR Problem Nobody Talks About Until It Is Too Late

Ask any ADC developer what their drug-to-antibody ratio is and they willgive you a number. Usually somewhere between 2 and 4, the range that mostapproved ADCs sit in. What that number does not tell you is the distributionunderneath it.

The drug-to-antibody ratio (DAR) describes how many cytotoxic payload moleculesare attached to each antibody in the ADC. In conventional stochasticconjugation approaches, whether lysine-based or partially reducedcysteine-based, that number is an average across a mixture. The actual productis a population of species with DAR 0, DAR 2, DAR 4, DAR 6, and DAR 8. The DAR0 species carry no payload and is therapeutically inert. The DAR 8 species ishighly hydrophobic, prone to aggregation, clears rapidly from circulation, andhas a narrow therapeutic window.

The analytical complexity of this heterogeneity is significant. A 2025 studypublished in the Journal of Medicinal Chemistry noted that stochasticconjugation produces not only different DAR values but positional isomerswithin each DAR value, meaning two DAR 4 molecules may have their four drugmolecules attached at completely different positions on the antibody, withdifferent pharmacokinetic consequences. Hydrophobic interaction chromatography,the standard DAR characterisation method, can show a consistent average DARvalue across batches while masking substantive variation in the underlyingisomeric distribution. A separate 2023 study found that apparent DAR valuesdetermined by reversed-phase LC-MS could fluctuate by up to 0.6 DAR unitssimply from variation in sample preparation steps before analysis.

Site-specific conjugation technologies address the homogeneity problemdirectly. Platforms including ThioMab, GlycoConnect, AJICAP, and AbClick enablecontrolled attachment of payloads at defined antibody positions, producinghomogeneous ADC preparations with a narrow DAR distribution rather than aheterogeneous mixture. PatSnap's April 2026 ADC manufacturing technologyanalysis described the sector's trajectory as a structural shift fromheterogeneous stochastic conjugation to homogeneous site-specific processes.The shift is real and it is accelerating, but it introduces its own formulationand manufacturing complexity, and many programmes in clinical development todaystill use stochastic conjugation platforms.

For developers: DAR characterisation must go beyond average value reportingfrom the earliest stages of development. The isomeric distribution, therelative proportions of DAR species, and the stability of that distributionunder accelerated storage and temperature stress conditions are all criticalquality attributes that regulators will expect to see characterised at IND andrigorously controlled at BLA. Programmes that defer this work until latedevelopment create a data gap that cannot be retroactively filled withoutcostly additional studies.

Challenge 2: Linker Instability, the Hidden Source of Off-Target Toxicity

The therapeutic logic of an ADC depends on one thing above all else: the linkermust be stable in systemic circulation and labile at the target site. Prematurepayload release in circulation, before the antibody has reached its tumourtarget, is what produces the off-target toxicity that has ended more than oneADC clinical programme.

The clinical record here is instructive. Mylotarg (gemtuzumab ozogamicin) wasthe first approved ADC, withdrawn from the US market in 2010 followingpost-approval studies showing excess mortality from hepatotoxicity, andre-approved in 2017 at a lower dose and with revised dosing schedule. Theprimary issue was premature hydrolysis of its acid-labile hydrazone linkerunder physiological conditions, releasing free calicheamicin payload incirculation. The linker chemistry had not been adequately characterised forsystemic instability before clinical development.

Two decades later, linker stability characterisation is more sophisticated, butthe challenge has not disappeared. It has become more complex. The val-citpeptide linker used in several approved ADCs including brentuximab vedotin andenfortumab vedotin was designed for protease-mediated cleavage insidelysosomes, where cathepsin B is active. It is not designed for cleavage incirculation. However, val-cit linkers are susceptible to hydrolysis in mouseplasma through a carboxylesterase-mediated mechanism, which creates a specificproblem for preclinical development: mouse plasma stability data does notpredict human plasma stability for val-cit-linked ADCs. Programmes that designaround mouse PK data for a val-cit ADC are working with data that does not translate.

A 2024 study published in the Journal of Medicinal Chemistry described theexolinker approach, a redesigned cleavable peptide linker architecture thatrepositions the cleavable element at the exo position of thep-aminobenzylcarbamate moiety. The exolinker design incorporates hydrophilicglutamic acid to address the aggregation-promoting hydrophobicity ofconventional val-cit linkers, while demonstrating resistance tocarboxylesterase and human neutrophil elastase-mediated cleavage, meaning thepayload remains attached in the presence of enzymes that compromiseconventional linkers. In vivo testing showed reduced premature payload release,increased achievable DAR values, and avoidance of significant aggregation evenwith hydrophobic payloads.

For developers: species selection in preclinical pharmacokinetic studiesrequires deliberate design choices depending on linker chemistry. A 2025Journal of Medicinal Chemistry review of peptide linkers for ADCs noted thatcleavable peptide linkers dominate clinical-stage ADC design but facechallenges including limited mouse plasma stability and structural homogeneityconstraints. The choice of linker chemistry and the species-specific stabilitydata package submitted to regulators must be matched, or the preclinical datatells the wrong story about a molecule's circulation stability.

A 2025 PMC review specifically titled "Formulation Matters: The OverlookedEngine of Stability and Success in Antibody-Drug Conjugates" was directabout what is at stake: "formulation variability including buffercomposition, excipient choice, ionic strength, and lyophilization can directlyaffect payload release, linker cleavage kinetics, and antibodyconformation." The formulation is not a packaging decision made at the endof development. It is an active determinant of how the linker behaves instorage, in transit, and in the patient's bloodstream.

Challenge 3: Aggregation — The Problem That Compounds Everything Else

If DAR heterogeneity is an upstream problem and linker stability is asystemic one, aggregation is the problem that sits at the intersection of bothand compounds every other formulation challenge an ADC developer faces.

ADC aggregation occurs through multiple, often simultaneous mechanisms.Hydrophobic payloads, particularly maytansinoids, auristatins, and the newercamptothecin derivatives, drive intermolecular hydrophobic interactions betweenADC molecules in solution. High DAR values increase the hydrophobic surfacearea exposed to solvent. Suboptimal conjugation conditions, includingincomplete reduction of disulfide bonds before cysteine conjugation, canintroduce structural heterogeneity that creates nucleation sites foraggregation. Manufacturing stresses including stirring, freeze-thaw cycles, andtemperature excursions accelerate the process. Even the excipient system,specifically the choice of surfactant, buffer, and stabiliser, determines howeffectively these aggregation drivers are controlled.

The clinical consequences are twofold. Aggregated ADC species have alteredpharmacokinetics, reduced target binding, and can trigger immunogenic responsesin patients. At the same time, aggregation during manufacturing is abatch-release failure that cannot be recovered, unlike some other qualitydeviations. An aggregated batch is either rejected or, in some cases, proceedswith a documented elevated aggregation level that requires regulatoryjustification. Neither outcome is acceptable in commercial supply.

The formulation response to ADC aggregation involves three parallel strategies:excipient optimisation, process design, and container closure selection. Onexcipient optimisation, the review published in PMC in 2026 confirmed thatlyophilisation with glass-forming matrices, specifically sucrose ortrehalose-based systems, combined with surfactants such as polysorbate 20 orpolysorbate 80 at defined concentrations, significantly enhances aggregationresistance during frozen storage and reconstitution. Buffer composition,particularly the choice of histidine versus phosphate versus citrate buffers,meaningfully affects conformational stability and aggregation onsettemperature. Ionic strength modulates electrostatic intermolecular repulsionand must be calibrated against viscosity requirements, particularly forprogrammes targeting subcutaneous delivery.

Subcutaneous ADC formulation is an area of rapidly growing clinical interestthat places concentrated aggregation pressure on formulation scientists.Sacituzumab govitecan and related compounds have been investigated forsubcutaneous administration, but as the review in Antibody Therapeutics (OxfordAcademic, 2025) documented, subcutaneous delivery requires high proteinconcentrations in small injection volumes, and at those concentrations, proteinaggregation and precipitation become the limiting constraints. The ADC'shydrophobic payload chemistry makes achieving the required concentrationwithout aggregation substantially harder than it is for unconjugatedantibodies.

Challenge 4: HPAPI Containment — The Safety Imperative That ShapesEverything Downstream

This challenge operates in a different register from the previous three.DAR control, linker stability, and aggregation are molecular problems thatformulation scientists solve at the bench and in process development. HPAPIcontainment is an infrastructure problem that shapes which facilities canhandle an ADC at all, what the capital costs of doing so are, and how theentire supply chain is designed.

ADC payloads are highly potent active pharmaceutical ingredients, a category ofcompounds defined by their extreme biological activity at low concentrations.Occupational exposure limits for ADC payloads are typically measured innanograms per cubic meter, meaning that a concentration of drug-containingaerosol too small to be visible by any conventional method is already at thelimit of what a worker can safely inhale. Auristatins, maytansinoids, andcamptothecin derivatives, the three dominant payload classes in clinical ADCdevelopment, all sit in this category.

The infrastructure required to handle these compounds at manufacturing scale isspecialised and expensive. High-containment suites must maintain negativepressure differential relative to surrounding areas, use continuous HVACfiltration systems capable of capturing particles at nanogram concentrations,and require closed-system processing equipment that eliminates operatorexposure pathways during conjugation, purification, formulation, andfill-finish. Technavio's 2026 ADC market analysis noted that handling HPAPIsrequires specialised high-containment facility infrastructure compliant withOccupational Exposure Band standards, and that this infrastructure increasesoperational costs by over 40% compared with standard pharmaceuticalmanufacturing suites. The cost and lead time of building or qualifyingHPAPI-capable facilities is one of the primary reasons that ADC developersincreasingly rely on CDMOs rather than building proprietary manufacturinginfrastructure.

The supply chain implications extend beyond the manufacturing site. ADC drugproducts require cold chain management from fill-finish through distribution toinfusion, most commonly at 2 to 8 degrees Celsius. Cold chain failures duringdistribution introduce temperature stress that accelerates aggregation andlinker degradation. Container closure systems for ADC vials require validationagainst the specific physicochemical properties of the ADC formulation,including compatibility testing between the stopper materials and theformulation excipients at the relevant storage temperature. A stopper that iscompatible with a standard antibody formulation may not be compatible with thesurfactant system required for an ADC with a hydrophobic payload at high concentration.

Challenge 5: Analytical Characterisation — Where Regulatory ExpectationsHave Moved Ahead of Industry Practice

The fifth challenge is not a manufacturing challenge or a stabilitychallenge. It is a characterisation challenge, specifically the gap betweenwhat many ADC programmes have historically provided as release andcharacterisation data and what regulatory agencies in the US, EU, andincreasingly Asia-Pacific now expect to see.

The analytical complexity of ADC characterisation reflects the molecule'sstructural complexity. A complete characterisation package for an ADC mustaddress the antibody component (including glycosylation profile, chargeheterogeneity, and structure), the linker-payload component (including theintegrity, purity, and stoichiometry of the attached linker-payload construct),the conjugate itself (including DAR distribution, positional isomer profile,unconjugated antibody content, free payload content, and high molecular weightaggregates), and the formulated drug product (including particle sizedistribution, subvisible particle counts, reconstitution behaviour forlyophilised products, and extractables and leachables from container closure).A 2026 PMC review noted that multi-attribute techniques based on massspectrometry, real-time kinetic modelling, and AI-driven design are redefininghow chemical stability and formulation outcomes are characterised andpredicted.

The FDA's guidance on analytical procedures for ADCs, reinforced by the EMA'sevolving expectations on ADC quality dossiers, has made the analytical releasespecification setting for DAR, aggregates, and free payload content aregulatory interaction point that earlier-stage programmes frequentlyunderestimate. Programmes that have characterised their ADC primarily byaverage DAR and basic purity assessment are at increasing risk of receivinganalytical-related questions during IND review or, worse, a clinical holdpending provision of additional characterisation data.

The analytical challenge is also a scale-up challenge. Characterisation methodsvalidated at laboratory scale must be re-validated at manufacturing scale, andthe behaviour of ADC samples in analytical workflows is sensitive to processingsteps. The 2023 study on DAR measurement variability found that apparent DARvalues could fluctuate by up to 0.6 DAR units from variation in sampledesalting and reduction steps before analysis, meaning that the analyticalmethod itself must be tightly controlled if DAR characterisation data is to bereproducible across batches and across sites.

The Through Line

Five challenges. Five distinct areas of failure. But they share a commonthread that every ADC developer working in 2026 needs to internalise.

ADC formulation is not a late-stage activity. The decisions that determine DARdistribution are made at conjugation chemistry design stage. The decisions thatdetermine linker stability in circulation are made when species are selectedfor preclinical PK studies. The decisions that determine aggregation behaviourat commercial concentration are made when early excipient screening is done orskipped. The decisions that determine whether a facility can even handle themolecule are made when the payload class is selected.

The ADC market is valued at approximately $8 billion in 2026 and is growing ata CAGR of 14.4%, projected to reach $20.56 billion by 2033. The pipeline has614 clinical candidates. The difference between the programmes thatsuccessfully navigate the formulation gauntlet and the ones that do not willnot be determined by the potency of their payload or the selectivity of theirantibody.

It will be determined by how early and how rigorously their formulation workwas done.

Sources: Antibody Therapeutics (Oxford Academic, 2025) "Fundamental properties and principal areas of focus in antibody-drug conjugates formulation development"; PMC "Formulation Matters: The Overlooked Engine of Stability and Success in Antibody-Drug Conjugates" (2026)

‍