Introduction
Antibody–drug conjugates (ADCs) combine the tumor-targeting specificity of monoclonal antibodies with the potency of cytotoxic payloads via a chemical linker. As of mid-2026, more than 17 ADCs have received regulatory approval globally 3, and the field has transitioned decisively from first-generation constructs toward next-generation platforms distinguished by site-specific conjugation, diverse payload classes, sophisticated linker architectures, and an expanding target landscape. This narrative review synthesizes the major design innovations, clinical advances, and safety considerations relevant to oncologists, hematologists, and translational medicine specialists from 2020 through July 2026.
From Conventional to Next-Generation ADC Design
Early ADCs suffered from several structural liabilities: random conjugation at lysine or cysteine residues produced heterogeneous drug-to-antibody ratio (DAR) distributions (typically 0–8), inconsistent pharmacokinetics (PK), and unpredictable safety profiles 2. Next-generation design addresses these limitations through four integrated axes.
Site-specific conjugation replaces random chemistry with engineered approaches, including incorporation of non-canonical amino acids (ncAAs) such as para-acetylphenylalanine (pAF), engineered cysteine residues (e.g., κK183C and K290C in PF-06804103), and enzymatic tags, yielding homogeneous, defined DAR products 67. ARX788 (HER2-directed, oxime-bond conjugation to pAF, DAR 2) exemplifies how site-specific methods achieve overlapping PK profiles for intact ADC and total antibody, indicating minimal free-drug release in vivo 6.
DAR optimization has shifted toward DAR 2–4 for improved stability and DAR 8 for maximized payload delivery when linker hydrophilicity is sufficient 21. Higher DARs increase hydrophobicity and hepatic clearance unless counteracted by hydrophilic spacers such as polyethylene glycol (PEG) or charged residues 34.
Hydrophilic linker engineering reduces aggregation propensity, extends circulation half-life, and improves solid-tumor penetration 3. In the 2026 phosphoramidate self-immolative linker platform, linker 11 (di-alanine with free carboxylate) conferred the most hydrophilic DAR 8 SN38-ADC tested, directly improving PK over more hydrophobic variants 1.
Bystander-effect optimization through traceless cleavable linkers allows released payload to diffuse into antigen-negative neighboring tumor cells, a mechanistic advantage in heterogeneous solid tumors 13.
Table 1. Major Next-Generation ADC Design Features and Their Clinical Rationale
| Design Feature | First-Generation Approach | Next-Generation Innovation | Clinical Rationale |
|---|---|---|---|
| Conjugation chemistry | Random lysine/cysteine conjugation | Site-specific (ncAA, engineered Cys, enzymatic) | Homogeneous DAR; predictable PK and safety |
| Drug-to-antibody ratio (DAR) | Heterogeneous (0–8) | Fixed (DAR 2, 4, or 8) | Reduced variability; tunable hydrophobicity |
| Linker hydrophilicity | Hydrophobic linkers | PEG spacers, charged residues, phosphoramidate scaffolds | Reduced aggregation; improved circulation half-life |
| Payload release | Linker-attached remnant on released payload | Traceless intracellular release (self-immolative) | Enhanced bystander effect; reduced off-target toxicity |
| Activation mechanism | Constitutive cleavage | Conditional (tumor microenvironment proteases, pH, reductase) | Selectivity for tumor vs. normal tissue |
| Antibody format | Monospecific IgG | Bispecific, Probody, biparatopic | Address antigen heterogeneity; improve internalization |
Payload Innovation
Three established payload classes—microtubule inhibitors, DNA-damaging agents, and topoisomerase I (Topo1) inhibitors—continue to anchor clinical ADC development, while emerging non-cytotoxic and immunostimulatory modalities are advancing preclinically 21.
Topoisomerase I inhibitors (e.g., DXd in trastuzumab deruxtecan, SN-38 in sacituzumab govitecan) have become the dominant payload class for solid tumors, offering single-digit nanomolar IC50 activity, moderate hydrophilicity enabling high DAR loads, and traceless release enabling bystander effects. In DESTINY-Breast03, trastuzumab deruxtecan (T-DXd; Enhertu) produced a median progression-free survival (PFS) of 29.0 months versus 7.2 months for trastuzumab emtansine (T-DM1), and improved median overall survival (OS) to 52.6 versus 42.7 months 10.
DNA-damaging agents (calicheamicin, pyrrolobenzodiazepines [PBD], duocarmycin) exhibit picomolar potency but carry significant hydrophobicity and off-target toxicity concerns 21. Trastuzumab duocarmazine demonstrated ocular toxicity in clinical testing, limiting its development in some programs 10.
Microtubule inhibitors (MMAE, MMAF, DM1, DM4) remain widely used in both hematologic and solid-tumor ADCs (e.g., brentuximab vedotin, polatuzumab vedotin, enfortumab vedotin), with peripheral neuropathy as the characteristic cumulative toxicity 25.
Emerging payloads include STING and TLR agonists for immunostimulatory ADCs (ISACs), PROTACs and LYTACs for degrader-antibody conjugates (DACs), and dual-payload constructs coupling mechanistically distinct cytotoxins 21. The dual-payload concept is particularly compelling given cross-payload resistance: TROPION-PanTumor01 data indicate that patients progressing on one Topo1i ADC show limited benefit from subsequent Topo1i ADCs, motivating mechanistic diversity 20. KH815, a TROP2-targeting ADC delivering a Topo1i and RNA polymerase II inhibitor, has entered clinical evaluation (NCT06885450) 20.
A 2026 Nature Communications study introduced phosphoramidate-based self-immolative linkers enabling traceless delivery of diverse hydroxy-containing payloads—including ganetespib (HSP90 inhibitor), paclitaxel, gemcitabine, and BAY-2402234 (DHODH inhibitor)—to clinically relevant antibodies at homogeneous DAR 8, demonstrating payload diversification beyond classical topoisomerase and tubulin targets 1.
Table 2. Representative Payload and Linker Classes in Next-Generation ADCs
| Payload Class | Representative Agents | Mechanism | Approximate Potency | Preferred Linker Type | Key Toxicity |
|---|---|---|---|---|---|
| Topoisomerase I inhibitors | DXd, SN-38 | Topo I inhibition | IC50 0.1–1 nM | Cleavable (cathepsin [DXd] or pH-sensitive [SN-38]) | ILD/pneumonitis, myelosuppression |
| Microtubule inhibitors | MMAE, MMAF, DM1, DM4 | Tubulin disruption | IC50 1–100 pM | Cleavable (protease, disulfide) | Peripheral neuropathy, cytopenias |
| DNA-damaging agents | Calicheamicin, PBD, Duocarmycin | DNA alkylation/crosslinking | IC50 0.1–10 pM | Acid-labile hydrazone (calicheamicin); cleavable dipeptide (PBD) | Myelosuppression, hepatotoxicity |
| Immunomodulatory payloads | STING agonists, TLR agonists | Innate immune activation | Variable | Cleavable (pH/protease) | Systemic immune activation (investigational) |
| Protein degraders | PROTACs, LYTACs | Ubiquitin-proteasome degradation | Variable | Cleavable (intracellular) | Under evaluation (preclinical) |
| Dual payloads | MMAE+MMAF, Topo1i+RNA Pol II inhibitor | Complementary mechanisms | Variable | Branched/orthogonal linkers | Under evaluation (early clinical) |
Linker Classes:
| Linker Type | Trigger | Serum Stability | Bystander Effect | Example ADCs |
|---|---|---|---|---|
| Non-cleavable (thioether) | Lysosomal antibody degradation | High | Low | T-DM1 (Kadcyla) |
| Protease-cleavable (dipeptide: Val-Cit, Ala-Ala) | Cathepsin B | Moderate–high | Moderate–high | Brentuximab vedotin, PF-06804103 |
| β-Glucuronidase-cleavable | Tumor microenvironment glucuronidase | High | High | Investigational |
| Disulfide (reductive) | Intracellular glutathione | Moderate | Moderate | Investigational |
| Phosphoramidate self-immolative | Esterase/protease/reductase | High (>50% intact ≥7 days) | High (traceless) | Investigational (saci-3-SN38, tras-15-DXd) |
| pH-sensitive | Acidic endosome/lysosome | Low–moderate | Moderate | Investigational |
The Expanding Target Landscape
The ADC target space has broadened substantially. Validated solid-tumor targets include HER2 (most mature; multiple approved ADCs across breast, gastric, and NSCLC), TROP2 (sacituzumab govitecan in triple-negative breast cancer [TNBC] and urothelial carcinoma; datopotamab deruxtecan approved January 2025 in HR+/HER2− breast cancer 14 and June 2025 in EGFR-mutated NSCLC 15), Nectin-4 (enfortumab vedotin), and FRα (mirvetuximab soravtansine approved March 2024 in platinum-resistant ovarian cancer, OS HR 0.67, p=0.0046) 16. c-Met was validated by the FDA accelerated approval of telisotuzumab vedotin (Emrelis) in May 2025 for high c-Met-overexpressing non-squamous NSCLC (ORR 35%, median DOR 7.2 months) 12.
Emerging late-stage solid-tumor targets include CLDN18.2 (multiple Phase III programs in gastric/oesophageal cancers), B7-H3 (ifinatamab deruxtecan showing dose-dependent activity in extensive-stage SCLC including CNS responses), HER3 (patritumab deruxtecan with intracranial activity in breast cancer and NSCLC brain metastases), and EGFR/MET dual-target bispecific constructs 1011.
In hematologic malignancies, CD30 (brentuximab vedotin), CD79b (polatuzumab vedotin), CD19 (loncastuximab tesirine), CD22 (inotuzumab ozogamicin), CD33 (gemtuzumab ozogamicin), and BCMA (belantamab mafodotin) constitute the validated core. ROR1 (zilovertamab vedotin, Phase III in DLBCL) and CD123/IL3RA (pivekimab sunirine, FDA approved June 2026 for BPDCN) illustrate the expanding hematologic target landscape 11.
Table 3. Key ADC Targets and Representative Agents (Global, 2020–2026)
| Target | Tumor Type | Representative ADC | Payload | Regulatory Status | Key Evidence |
|---|---|---|---|---|---|
| HER2 | Breast, gastric, NSCLC, pan-tumor HER2+ | Trastuzumab deruxtecan (Enhertu) | DXd (Topo1i) | FDA/EMA/NMPA approved | DESTINY-Breast03: PFS 29.0 vs 7.2 mo (vs T-DM1) 10 |
| TROP2 | HR+/HER2− breast cancer | Datopotamab deruxtecan (Datroway) | DXd (Topo1i) | FDA approved Jan 2025 | TROPION-Breast01: PFS HR 0.63 14 |
| TROP2 | EGFR-mutated NSCLC | Datopotamab deruxtecan (Datroway) | DXd (Topo1i) | FDA accelerated approval Jun 2025 | ORR 45% 15 |
| Nectin-4 | Urothelial carcinoma | Enfortumab vedotin (Padcev) | MMAE | FDA/EMA/NMPA approved | — |
| FRα | Platinum-resistant ovarian cancer | Mirvetuximab soravtansine (Elahere) | DM4 | FDA approved Mar 2024 | MIRASOL: OS HR 0.67 16 |
| c-Met | High c-Met+ non-squamous NSCLC | Telisotuzumab vedotin (Emrelis) | MMAE | FDA accelerated approval May 2025 | LUMINOSITY: ORR 35% 12 |
| BCMA | Relapsed/refractory multiple myeloma | Belantamab mafodotin (Blenrep) | MMAF | FDA approved Oct 2025 (BVd combination) | DREAMM-7: PFS 31.3 vs 10.4 mo 13 |
| CD79b | DLBCL | Polatuzumab vedotin (Polivy) | MMAE | FDA/EMA approved | POLARIX 5-yr: PFS 64.2% vs 59.1% 10 |
| CD30 | PTCL, Hodgkin lymphoma | Brentuximab vedotin (Adcetris) | MMAE | FDA/EMA approved | ECHELON-2: 5-yr OS 68.7% vs 60.3% 10 |
| CLDN18.2 | Gastric, oesophageal | Multiple (AZD-0901, IBI-343, LM-302) | Various | Phase III | Multiple ongoing trials 11 |
| B7-H3 | SCLC, solid tumors | Ifinatamab deruxtecan (DS-7300a) | DXd (Topo1i) | Phase III | CNS responses in SCLC 10 |
| HER3 | NSCLC, breast cancer | Patritumab deruxtecan (HER3-DXd) | DXd (Topo1i) | Phase III | Intracranial activity in BM 10 |
| ROR1 | DLBCL | Zilovertamab vedotin | MMAE | Phase III | — 11 |
| HER2 (pan-tumor) | HER2 IHC 3+ solid tumors | Trastuzumab deruxtecan (Enhertu) | DXd | FDA accelerated approval Apr 2024 | DESTINY-PanTumor02: ORR 51.4% 18 |
Clinical Advances: Selected Highlights, 2020–2026
Table 4. Selected Major Clinical Advances and Regulatory Approvals, 2020–2026
| Milestone | Agent / Trial | Setting | Key Efficacy Finding | Key Safety Signal | Regulatory Outcome |
|---|---|---|---|---|---|
| HER2+ mBC 2L standard shift | T-DXd / DESTINY-Breast03 | Post-T-DM1 HER2+ mBC | OS 52.6 vs 42.7 mo; PFS 29.0 vs 7.2 mo | ILD 16.7% | FDA approved 10 |
| T-DXd post-T-DM1 confirmation | T-DXd / DESTINY-Breast02 | Post-T-DM1 HER2+ mBC | PFS 17.8 vs 6.9 mo; OS 39.2 vs 26.5 mo | ILD 10.4% (grade 5: 1.7%) | FDA approved 10 |
| T-DXd 1L mBC | T-DXd + pertuzumab / DESTINY-Breast09 | 1L HER2+ mBC | PFS 40.7 vs 26.9 mo (vs THP) | ILD 12.1%, incl. Grade 5 | Regulatory review 10 |
| CNS activity | T-DXd / DESTINY-Breast12 | HER2+ mBC with brain metastases | 12-mo PFS 61.6%; CNS ORR 62.3% (active BM) | Fatal ILD reported | Supports CNS use 10 |
| Pan-tumor HER2+ approval | T-DXd / DESTINY-PanTumor02 | HER2 IHC 3+ solid tumors | ORR 51.4%; median DOR 19.4 mo | ILD/pneumonitis Boxed Warning | FDA accelerated approval Apr 2024 18 |
| FRα ovarian cancer | Mirvetuximab soravtansine / MIRASOL | Platinum-resistant ovarian cancer | OS HR 0.67 (p=0.0046); ORR 42% vs 16% | Ocular toxicity (Boxed Warning) | FDA approved Mar 2024 16 |
| TROP2 HR+/HER2− breast cancer | Datopotamab deruxtecan / TROPION-Breast01 | Post-endocrine + post-chemo mBC | PFS HR 0.63 (p<0.0001); ORR 36% vs 23% | Stomatitis, keratitis | FDA approved Jan 2025 14 |
| c-Met NSCLC | Telisotuzumab vedotin / LUMINOSITY | High c-Met+ non-squamous NSCLC | ORR 35%; median DOR 7.2 mo | Peripheral neuropathy, edema | FDA accelerated approval May 2025 12 |
| BCMA myeloma combination rescue | Belantamab mafodotin + Vd / DREAMM-7 | RRMM ≥2 prior lines | PFS 31.3 vs 10.4 mo vs DVd; OS HR 0.49 | Ocular toxicity 92% (Grade 3/4 77%) | FDA approved Oct 2025 13 |
| Frontline DLBCL | Polatuzumab vedotin + R-CHP / POLARIX | Untreated DLBCL | 5-yr PFS 64.2% vs 59.1% (vs R-CHOP) | Neuropathy, cytopenias | Established standard 10 |
| Frontline PTCL | Brentuximab vedotin + CHP / ECHELON-2 | CD30+ PTCL | 5-yr OS 68.7% vs 60.3% | Neuropathy (manageable) | Established standard 10 |
A critical practical lesson from this dataset is that belantamab mafodotin illustrates the importance of context of use: DREAMM-3 failed to meet its primary PFS endpoint as monotherapy versus Pom/Dex, yet DREAMM-7 (combination with bortezomib-dexamethasone) demonstrated highly significant PFS benefit (HR 0.31) and OS benefit (HR 0.49), prompting FDA approval in October 2025 13. Conversely, the EMA did not renew Blenrep's conditional marketing authorization in February 2024 for the heavily pretreated monotherapy indication, reflecting the challenging single-agent benefit-risk profile 17. These contrasting outcomes underscore that ADC value is not platform-intrinsic but depends critically on treatment line, combination architecture, and toxicity management infrastructure.
Safety and Toxicity Management
Platform-specific toxicity profiles are a defining clinical reality for ADCs.
Interstitial lung disease/pneumonitis (ILD) is the hallmark concern for Topo1i ADC platforms. In T-DXd studies, ILD occurs in 10–17% of patients, with fatal cases reported across DESTINY-Breast02, DESTINY-Breast09, and DESTINY-Breast12 22. The structured "Five S" algorithm (Screen, Scan, Synergy, Suspend, Steroids) guides management: Grade 1 asymptomatic ILD warrants dose interruption and corticosteroids (≥0.5 mg/kg/day prednisolone), while Grade 2 or greater mandates permanent discontinuation 22.
Ocular toxicity is the defining adverse event for belantamab mafodotin (keratopathy in >70% of patients, requiring dose modification in 83%) 2324 and is relevant for mirvetuximab soravtansine (blurred vision, keratopathy with Boxed Warning) 16 and datopotamab deruxtecan (keratitis in ≥20%) 14. Baseline and periodic ophthalmologic assessment using the Keratopathy and Visual Acuity (KVA) scale is mandated for belantamab-containing regimens 23.
Peripheral neuropathy is cumulative and class-characteristic for MMAE/MMAF-based ADCs (brentuximab vedotin, enfortumab vedotin, polatuzumab vedotin, zilovertamab vedotin). ADCETRIS prescribing information specifies dose interruption and reduction to 1.2 mg/kg every 3 weeks for Grade 2–3 neuropathy, with permanent discontinuation for Grade 4 25.
Myelosuppression is common across ADC families and becomes the principal dose-limiting toxicity driver in combination regimens. Clinicians should anticipate additive cytopenias when ADCs are combined with chemotherapy or immunomodulatory agents 10.
Emerging Platforms and Future Directions
Beyond approved agents, next-generation ADC formats under active investigation include bispecific ADCs (targeting EGFR+MET, HER2+PD-L1, TROP2+PD-L1), Probody-drug conjugates (protease-activated in the tumor microenvironment), ISACs (STING/TLR agonist payloads), DACs (PROTAC-based), and dual-payload constructs 21. The phosphoramidate self-immolative linker platform represents a significant chemistry advance, enabling traceless delivery of diverse hydroxy-containing payloads (paclitaxel, gemcitabine, ganetespib, and others) at homogeneous DAR 8 with superior in vitro selectivity and favorable preclinical PK compared to approved comparators such as sacituzumab govitecan and trastuzumab deruxtecan 1. Dual-payload ADC KH815 has entered early-phase clinical evaluation 20, and the field anticipates that mechanistic diversity within a single conjugate will be central to overcoming cross-payload resistance.
Conclusions and Clinical Implications
From 2020 to July 2026, next-generation ADCs have redefined standards across HER2-positive breast cancer, platinum-resistant ovarian cancer, NSCLC, multiple myeloma, and aggressive lymphomas, while progressively moving from salvage monotherapy into frontline combination use and CNS-active settings. The convergence of site-specific conjugation, hydrophilic linker engineering, traceless self-immolative release, and payload diversification addresses the fundamental limitations of first-generation constructs. Clinically, ADC success now depends equally on target selection, treatment line, combination design, and toxicity co-management expertise. Cross-trial comparisons remain hazardous given population heterogeneity, control arm differences, and endpoint maturity. Ongoing development of dual-payload, conditionally activated, and immunostimulatory ADCs promises further expansion of the therapeutic arsenal, though comprehensive clinical safety frameworks for these modalities are still maturing.