Antibody-Drug Conjugate

An antibody-drug conjugate (ADC) is a tripartite biopharmaceutical therapeutic that covalently joins a tumor-targeting monoclonal antibody (mAb) to a potent cytotoxic small-molecule payload via a specialized chemical linker. Conceived as a molecular realization of Paul Ehrlich's early twentieth-century concept of the Zauberkugel ("magic bullet"), an ADC combines the nanomolar antigen specificity and multi-week FcRn-mediated serum half-life of an immunoglobulin with a small-molecule warhead whose intrinsic cytotoxicity (IC₅₀ ≈ 10⁻¹¹ to 10⁻⁹ M) is 100- to 1,000-fold too potent to be administered systemically as an unconjugated free drug.

Following intravenous infusion, the intact ADC circulates in the bloodstream, extravasates into the tumor interstitium, and binds its cognate cell-surface antigen. Antigen engagement triggers clathrin- or caveolae-mediated receptor-mediated endocytosis, trafficking the ADC through early and late endosomes into the acidic, protease-rich lysosome. Inside the lysosome, chemical or enzymatic cleavage of the linker—or complete proteolytic digestion of the antibody scaffold—liberates the active warhead to disrupt microtubules or damage DNA, inducing apoptotic cell death.

Target Antigen Selection and Conjugation Chemistry

Optimal ADC performance depends on balancing four quantitative design parameters across the antibody, conjugation site, linker, and payload:

1. Target Antigen and Antibody: The target antigen should be abundantly expressed on the apical or extracellular tumor cell membrane (typically >10⁴ to 10⁵ copies per cell), show restricted expression on vital normal tissues, and internalize efficiently upon antibody cross-linking rather than shedding into plasma. Clinically validated ADC targets include HER2 (ERBB2), Trop-2 (TACSTD2), Nectin-4, Folate Receptor α (FOLR1), Tissue Factor, CD30, CD33, CD22, CD19, CD79b, and BCMA. Humanized or fully human IgG1 is the predominant backbone.
2. Conjugation Chemistry and Drug-to-Antibody Ratio (DAR): The Drug-to-Antibody Ratio (DAR) represents the average number of warhead molecules attached per antibody molecule. Three conjugation strategies govern DAR homogeneity:

• Stochastic Lysine Conjugation: Reacting activated N-hydroxysuccinimide (NHS) esters with solvent-accessible surface lysine ε-amine groups (~80 lysines exist per IgG, of which ~20–30 are sterically accessible) produces a heterogeneous Poisson-like distribution of DAR 0 to 8 species with an average DAR ≈ 3.5 (used in ado-trastuzumab emtansine [T-DM1], gemtuzumab ozogamicin, and inotuzumab ozogamicin).
• Interchain Disulfide Cysteine Conjugation: Partial reduction of the four inter-chain disulfide bonds of IgG1 with tris(2-carboxyethyl)phosphine (TCEP) exposes up to eight reactive cysteine sulfhydryl (–SH) groups, which react with maleimide-functionalized linkers. Partial alkylation yields even-numbered species (DAR 0, 2, 4, 6, 8) with an average DAR ≈ 4 (as in brentuximab vedotin, polatuzumab vedotin, and enfortumab vedotin), whereas full reduction of all four interchain disulfides coupled with hydrophilic linker masking yields a homogeneous DAR ≈ 7.6 to 8.0 (as in trastuzumab deruxtecan [T-DXd] and sacituzumab govitecan).
• Site-Specific Conjugation: Engineering unpaired cysteines (THIOMABs), incorporating non-canonical amino acids (such as para-acetylphenylalanine for oxime ligation), or using enzymatic glycan remodeling / bacterial transglutaminase produces a uniform DAR = 2 or 4. Without hydrophilic masking (such as PEGylation or polysarcosine), unmasked hydrophobic species with DAR > 6 aggregate in plasma and undergo rapid non-specific uptake by hepatic Kupffer cells and sinusoidal endothelial cells, accelerating systemic clearance and causing hepatotoxicity without improving tumor delivery.

The chemical linker must maintain stability in human plasma (preventing premature systemic deconjugation that causes off-target neutropenia, neuropathy, or thrombocytopenia) while rapidly releasing the active payload once inside the tumor lysosome:

Non-Cleavable Linkers

Non-cleavable linkers—such as the maleimidomethyl cyclohexane-1-carboxylate (MCC) thioether linker in ado-trastuzumab emtansine (T-DM1) or maleimidocaproyl (mc) linkers conjugated to MMAF (as in belantamab mafodotin)—contain no chemical or enzymatic cleavage trigger. After endocytosis, the entire monoclonal antibody must be proteolytically digested down to individual amino acids inside the lysosome. Consequently, the liberated active species retains the conjugated amino acid residue appended to the linker and payload (for T-DM1, positively charged lysine–MCC–DM1). Because this zwitterionic/charged catabolite cannot passively diffuse across the hydrophobic lipid bilayer of the plasma membrane, non-cleavable ADCs exhibit high plasma stability and low systemic off-target toxicity, but cannot kill neighboring antigen-negative tumor cells.

Chemically and Enzymatically Cleavable Linkers

Cleavable linkers exploit biochemical differences between extracellular blood plasma (pH 7.4, low glutathione, protease inhibitors present) and the intracellular tumor endolysosome:

1. Acid-Labile Linkers: Contain a hydrazone (in gemtuzumab ozogamicin and inotuzumab ozogamicin) or a pH-sensitive benzyl carbonate (CL2A) bond (in sacituzumab govitecan) that remains relatively stable at blood pH 7.4 but undergoes acid-catalyzed hydrolysis inside endosomes (pH 5.5–6.2) and lysosomes (pH 4.5–5.0).
2. Reducible Disulfide Linkers: Exploit the ~1,000-fold concentration gradient between low extracellular plasma glutathione (GSH ≈ 2–20 μM, oxidizing environment) and high intracellular cytoplasmic/endosomal thiol levels (GSH ≈ 1–10 mM, reducing environment), exemplified by the hindered disulfide linker in mirvetuximab soravtansine (DM4 payload).
3. Protease-Cleavable Peptide Linkers: Incorporate a short oligopeptide substrate selectively recognized by lysosomal cysteine proteases such as cathepsin B and legumain, which are sequestered inside lysosomes (and inactivated at pH 7.4 in plasma by cystatins). Two canonical designs dominate clinical practice:

• The valine–citrulline (Val–Cit) dipeptide coupled to a self-immolative para-aminobenzyloxycarbonyl (PABC) spacer (used in brentuximab vedotin, polatuzumab vedotin, enfortumab vedotin, and tisotumab vedotin): cathepsin B cleaves the amide bond C-terminal to citrulline, triggering a spontaneous 1,6-elimination cascade across the PABC spacer (releasing CO₂ and aza-quinone methide) to liberate unmodified, neutral free MMAE.
• The hydrophilic glycine–glycine–phenylalanine–glycine (GGFG) tetrapeptide coupled to a compact self-immolative aminomethylene spacer (used in trastuzumab deruxtecan [T-DXd] and datopotamab deruxtecan), which reduces linker hydrophobicity so that eight DXd molecules (DAR = 8) can be attached without antibody aggregation.

Why Picomolar Payloads Are Required

Radiolabeled biodistribution studies in human patients demonstrate that following intravenous administration, only 0.01% to 0.1% of the injected ADC dose per gram of tissue (and <1%–2% of the total dose) localizes to solid tumors, with intracellular payload accumulation limited by receptor copy number and endocytosis rate. First-generation ADCs conjugated to conventional chemotherapeutics such as doxorubicin, methotrexate, or 5-fluorouracil (IC₅₀ ≈ 10⁻⁷ to 10⁻⁶ M) failed clinically because the intracellular drug concentration achieved inside tumor cells fell below the cytotoxic threshold. Modern ADCs therefore use warheads in three classes:

• Microtubule Inhibitors (IC₅₀ ≈ 10⁻¹¹ to 10⁻⁹ M): Synthetic dolastatin-10 analogues (auristatins: neutral membrane-permeable MMAE and charged C-terminal phenylalanine MMAF) and ansamycin macrolides (maytansinoids: DM1 and DM4), which bind the Vinca/maytansine domain of β-tubulin to arrest cells in G2/M phase.
• DNA-Cleaving and Crosslinking Agents (IC₅₀ ≈ 10⁻¹² to 10⁻¹⁰ M): Enediyne antibiotics (calicheamicin γ₁, which undergoes Bergman cycloaromatization in the DNA minor groove to abstract hydrogen atoms and cause double-strand breaks) and pyrrolobenzodiazepine (PBD) dimers (tesirine in loncastuximab tesirine, forming sequence-selective minor-groove interstrand DNA crosslinks).
• Topoisomerase I Inhibitors (IC₅₀ ≈ 10⁻⁹ to 10⁻⁸ M): Camptothecin derivatives SN-38 and deruxtecan (DXd), which stabilize the covalent Topoisomerase I–DNA cleavage complex, generating lethal replication-associated double-strand breaks.

The Bystander Killing Effect in Heterogeneous Solid Tumors

Solid tumors frequently exhibit intratumoral heterogeneity in antigen expression: HER2-positive cells may sit adjacent to HER2-low or HER2-negative tumor cells. When an ADC carrying a cleavable linker (such as GGFG or Val–Cit) releases an uncharged, lipophilic payload (DXd, SN-38, or MMAE), the free drug kills the target-positive cell and diffuses across the plasma membrane into the local tumor microenvironment to eradicate neighboring antigen-negative tumor cells—a phenomenon termed the bystander killing effect.

This mechanistic distinction explains the landmark clinical divergence between T-DM1 (non-cleavable MCC linker, DAR ≈ 3.5, charged Lys–MCC–DM1 catabolite, no bystander effect) and trastuzumab deruxtecan (T-DXd) (cleavable GGFG linker, DAR ≈ 8, membrane-permeable DXd payload with a short systemic half-life of ~1.4 hours to limit distant toxicity). In the Phase III DESTINY-Breast04 trial, T-DXd significantly prolonged progression-free and overall survival in patients with HER2-low metastatic breast cancer (immunohistochemistry IHC 1+ or IHC 2+/in situ hybridization negative)—a population constituting ~50% of all breast cancers in which T-DM1 and unconjugated trastuzumab had previously shown no clinical activity.

Why does maleimide–thiol cysteine conjugation undergo retro-Michael deconjugation in plasma?

Classical cysteine-conjugated ADCs form a uccinimidyl thioether bond via Michael addition of a cysteine thiol to a maleimide ring. In blood plasma, this thiosuccinimide linkage can undergo a slow reversible retro-Michael reaction (half-life ~5–7 days), transferring the maleimide–linker–payload onto the reactive Cys34 thiol of abundant human serum albumin (~600 μM). Second-generation self-hydrolyzing maleimides artificially trigger succinimide ring hydrolysis into an open-ring succinamic acid thioether, which is chemically resistant to retro-Michael deconjugation.

What causes ocular keratopathy and interstitial lung disease (ILD) with certain ADCs?

Most dose-limiting toxicities of ADCs are target-independent (off-target, on-payload) and driven by non-specific pinocytosis or FcγR-mediated uptake of intact ADC by normal cells, or premature systemic payload release. For example, MMAF- and DM4-conjugated ADCs frequently cause reversible microcystic corneal epithelial keratopathy, MMAE ADCs cause peripheral sensory neuropathy and neutropenia, and trastuzumab deruxtecan carries an ~10%–15% risk of interstitial lung disease (ILD) / pneumonitis thought to involve uptake of high-DAR IgG1 by alveolar macrophages.

Why can Topoisomerase I ADCs (T-DXd, sacituzumab govitecan) use a DAR of 8 whereas auristatin/PBD ADCs are restricted to DAR 2–4?

Topoisomerase I payloads (DXd, SN-38) have moderately lower intrinsic potency (IC₅₀ ~1–10 nM) than PBD dimers (picomolar) or auristatins, allowing higher cellular payload delivery at DAR = 8 without exceeding systemic tolerability. Combining a hydrophilic linker (GGFG in T-DXd or PEGylated CL2A in sacituzumab govitecan) with camptothecin derivatives prevents the hydrophobic aggregation that plagues DAR = 8 auristatin conjugates.

1. Beck, A., Goetsch, L., Dumontet, C., & Corvaïa, N. (2017). "Strategies and challenges for the next generation of antibody–drug conjugates." Nature Reviews Drug Discovery, 16(5), 315–337. https://doi.org/10.1038/nrd.2016.268
2. Ogitani, Y., Aida, T., Hagihara, K., et al. (2016). "DS-8201a, a novel HER2-targeting ADC with a novel DNA topoisomerase I inhibitor, demonstrates a promising antitumor efficacy with differentiation from T-DM1." Clinical Cancer Research, 22(20), 5097–5108. https://doi.org/10.1158/1078-0432.CCR-15-2873
3. Modi, S., Jacot, W., Yamashita, T., et al. (2022). "Trastuzumab deruxtecan in previously treated HER2-low advanced breast cancer." New England Journal of Medicine, 387(1), 9–20. https://doi.org/10.1056/NEJMoa2203690
4. Drago, J. Z., Modi, S., & Chandarlapaty, S. (2021). "Unlocking the potential of antibody–drug conjugates for cancer therapy." Nature Reviews Clinical Oncology, 18(6), 327–344. https://doi.org/10.1038/s41571-021-00470-8