ADC Target Identification for Resistant Tumors

Nature paper (15 July 26) demonstrates that acquired ADC resistance can be pharmacologically reversed in vivo by snapping two FDA-approved therapeutic antibodies together inside the body via bioorthogonal click chemistry, without antibody re-engineering. First posted on bioRxiv in Sept 2023, the paper required a long peer review: reviewers challenged whether the benefit arose from tumor-site click ligation vs bispecific-like complex formation in the bloodstream, and the authors answered with antigen-negative cell controls, pre-clicked biodistribution comparisons, and a site-specific conjugation dataset. 🔅 The platform exploits the inverse electron demand Diels-Alder reaction between trans-cyclooctene (TCO) and tetrazine (Tz). Each antibody or ADC is independently conjugated with one handle and administered sequentially; after the first antibody localizes to its tumor antigen, the second ligates with it in vivo, redirecting the ADC payload to tumor regions expressing low or no target for the ADC alone. The spatial rationale is hard data: HER2 and EGFR showed near-zero regional overlap in tumor specimens, with Dice similarity coefficient values of 0.12 and 0.26. EGFR-targeting panitumumab-TCO was paired with HER2 ADCs T-DXd or T-DM1 bearing Tz. Site-specific Fc glycan conjugation placed just 2-4 handles per antibody, reduced liver uptake versus random lysine conjugation (p = 0.0064, n = 4 mice), and achieved 24.93 ± 2.75 max %ID/mL tumor uptake at 120 hours versus 12.67 ± 1.72 for pre-clicked controls (p = 1.51 × 10⁻⁴). 🔆 In a HER2-low admixed breast cancer xenograft, a single 5 mg/kg dose of T-DXd click outperformed monotherapy and non-clicked controls (p = 0.0018, n = 10-14 mice; log-rank p = 1.87 × 10⁻⁷). In HER2-negative pancreatic xenografts (MIAPaCa-2), the advantage was larger still (no click vs click p < 1.10 × 10⁻³²³; survival log-rank p = 2.28 × 10⁻⁸, n = 8-10 mice). NCIN87 tumors failing T-DXd showed a 2.1-fold EGFR upregulation; switching 9 non-responders to panitumumab-TCO plus T-DXd-Tz click controlled tumors in 5. Among 6 trastuzumab-resistant BT474 T-DXd non-responders, pertuzumab-TCO plus T-DXd-Tz click suppressed growth in 5. Click ligation extended to antibody-TCO conjugates targeting PD-L1, PSMA, and VEGF. 👉Limitation: All efficacy data are from mouse xenografts only. Ligation occurs partly in the bloodstream, and enhanced internalization cannot be cleanly separated from antibody clustering or pharmacokinetic effects. Preclinical doses of 5-20 mg/kg correspond to 0.4-1.6 mg/kg human equivalents. DXd pulmonary toxicity was not evaluated in humanized models. 📑 Full text: https://lnkd.in/gGufizht 📑 Preprint: https://lnkd.in/gkCcanQ4 #AntibodyDrugConjugates #BioorthogonalChemistry #CancerTherapy

What if two antibodies could find each other and assemble a cancer drug — inside the body, after you've already dosed the patient? https://lnkd.in/g6jduuaJ Antibody–drug conjugates (ADCs) have reshaped oncology by delivering potent cytotoxic payloads straight to tumour cells. But they carry a built-in limitation: each ADC hunts for a single target antigen. In heterogeneous tumours — where that antigen is expressed unevenly, weakly, or not at all — targeting breaks down and resistance takes hold. A new preclinical study introduces a clever way around that ceiling: build the therapeutic construct in vivo, using click chemistry. Here's the idea: → Tag a therapeutic antibody with one bioorthogonal handle (trans-cyclooctene) and an ADC with its chemical partner (tetrazine). → Administer them sequentially. Once both are circulating, the two halves ligate to each other in the body — forming a functional antibody–ADC "click" construct at the tumour. → The result is a drug assembled on site, no extensive antibody re-engineering required. Why it matters: in models co-expressing HER2 and EGFR, this antibody–ADC click approach outperformed standard ADC monotherapy and simple antibody-plus-ADC combinations. Crucially, it showed activity in tumours with low, ultralow, negative, or heterogeneous HER2 expression — precisely the cases that are resistant to, or ineligible for, conventional HER2-directed ADCs. For those of us working in biomarker scoring and companion diagnostics, this is a compelling shift. So much of our effort goes into stratifying patients by expression thresholds. A modular platform that engages receptor biology across a spectrum of expression levels — and extends to other receptor pairs — could reframe how we think about eligibility, heterogeneity, and resistance altogether. The payload isn't the only thing that matters. Increasingly, it's the chemistry that decides where and how the payload comes together. #ADCs #AntibodyDrugConjugates #HER2 #Oncology #ClickChemistry #BioorthogonalChemistry #CancerResearch #PrecisionMedicine #CompanionDiagnostics #TargetedTherapy #TumorHeterogeneity Figure Courtesy: Nature. Patrícia M. R. Pereira Washington University School of Medicine, St Louis, MO, USA.

【CD44v9 as a novel target】  Antibody–drug conjugates (ADCs) have reshaped advanced breast cancer care—but today, benefit is still largely limited to HER2- and TROP2-directed therapies. Resistance remains inevitable for many patients. We need new, tumor-selective targets that persist in refractory disease. Our work identifies CD44v9 as one such target. CD44v9 is highly expressed in aggressive breast cancers and largely absent in normal tissues, including peripheral blood mononuclear cells. It is linked to stemness, redox balance, and therapy resistance—key drivers of recurrence. We developed a CD44v9-targeted ADC (anti-CD44v9–MMAF) that shows strong, specific binding, efficient internalization, potent antigen-dependent cytotoxicity, and dose-dependent tumor inhibition with minimal systemic toxicity. Importantly, CD44v9 expression persists in models resistant to trastuzumab deruxtecan, sacituzumab govitecan, and T-DM1—and these resistant tumors remain sensitive to CD44v9-directed therapy. CD44v9 may represent a clinically actionable, tumor-selective ADC target that extends the treatment continuum beyond HER2 and TROP2 for patients with advanced, refractory breast cancer. Collective outcome of teamwork by the Preclinical Core of the University of Hawai'i Cancer Center by Dr. Jangsoon Lee, Dr. Kyoji Tsuchikama (UT Houston), and Dr. Jennifer Maynard (UT Austin) https://lnkd.in/gxMjERaw

My article of the week. Those involved in oncology drug development and clinical practice are likely familiar with checkpoint inhibitors (CPIs) and their dramatic role in the battle against cancer. But, resistance to CPIs such as anti-PD-1 and anti-PD-L1 remains a significant challenge in the treatment of many solid tumours. Patients who fail CPI therapy have few alternative treatment options, highlighting a high unmet need. My article of the week explores GDF-15 (growth differentiation factor 15), a tumour-derived cytokine, as a potential driver of immune evasion and resistance to CPI therapies. Elevated GDF-15 levels are associated with reduced T-cell infiltration and impaired tumour immune response. Mechanism: The study identifies multiple mechanisms by which GDF-15 suppresses the immune response in the tumour microenvironment. GDF-15 affects it by: • Reducing T-cell recruitment and activation. • Modulating chemokine expression and immune surveillance pathways. • Inhibiting both innate and adaptive immune responses required for effective CPI treatment. Therapeutic Approach: The study investigated visugromab, a monoclonal antibody targeting GDF-15. Preclinical models demonstrated increased T-cell infiltration and enhanced anti-tumour activity when GDF-15 was neutralised. Combination therapy with CPI showed synergistic effects, overcoming immune suppression in resistant tumour models. Clinical Data: These findings translated into promising early clinical data. Preliminary clinical results in patients with refractory non-small cell lung cancer (NSCLC) and urothelial carcinoma showed: • Increased T-cell infiltration in tumour biopsies. • Durable responses in a subset of patients previously unresponsive to CPI therapies. This study highlights a new avenue to address CPI resistance by targeting GDF-15, that should be watched carefully #CheckpointInhibitors #CancerResearch #ImmunoOncology #DrugDevelopment #ClinicalInnovation

Rebuilding pancreatic cancer in a dish to decode and target the tumor-stroma alliance Pancreatic ductal adenocarcinoma (PDAC) remains among the deadliest cancers, with survival stubbornly low despite decades of therapeutic attempts. One major challenge is PDAC’s dense, fibroblast-rich stroma, which shields cancer cells and fuels aggressive evolution. A new study combines single-cell atlasing with patient-derived, stroma-integrated tumoroids to expose how tumor and stroma coevolve, and where that alliance can be broken. Key breakthroughs: ◾ Researchers integrated >670,000 single-cell transcriptomes from 200 PDAC samples to build the first high-resolution pancreatic cancer cell atlas, capturing conserved epithelial, fibroblast, and stromal states across patients. ◾ Using this atlas as a reference, they engineered modular “tumoroids”, 3D co-cultures combining cancer and cancer-associated fibroblast (CAF) populations that recapitulate ductal structures and desmoplastic stroma. ◾ Both atlas and tumoroids revealed a recurrent signaling hub centered on Syndecan-1 (SDC1), a CAF-responsive receptor whose high expression correlates with poor survival. Blocking SDC1 with an antibody-drug conjugate (ADC) disrupted cancer growth, extracellular matrix organization, and CAF communication. This work does more than identify a target, it delivers a systems-level dissection of PDAC. By integrating patient-derived atlases with tractable 3D models, the study bridges the divide between discovery and preclinical validation. The approach enables: ▪️ Mechanistic testing of tumor–stroma interactions that drive resistance and metastasis. ▪️ Benchmarking of model fidelity against patient data, de-risking translational gaps. ▪️ Actionable drug screening platforms that reflect in vivo heterogeneity, essential for advancing next-generation therapeutics, including ADCs, stromal modulators, and microenvironment-directed immunotherapies. This study exemplifies where oncology needs to go, beyond the cancer cell. As single-cell atlases mature and in vitro systems become physiologically faithful, we can now reconstruct the emergent ecosystems that sustain malignancy. The identification of SDC1 as a stromal-activated node underscores that many therapeutic answers in PDAC may lie not in new oncogenes, but in decoding, and dismantling, the communication circuits that protect them. Read the full study: https://lnkd.in/eWUryr2c #SingleCellBiology #TumorMicroenvironment #PancreaticCancer #DrugDevelopment #TranslationalResearch

Does high target expression lead to strong ADC response? Not always. Enfortumab vedotin (a NECTIN4-binding ADC) is one of the ADC success stories in urothelial carcinoma. But approx. 40% of patients never respond, and many who initially benefit, eventually encounter progression. A recent Cancer Cell study (2026, link below 👇 ) shows why "NECTIN4-high" tumors can still escape the drug. They found a resistant tumor population that is NECTIN4-high but uptake-defective and this is the escape mechanism:   1️⃣ Less ADC enters the cell: an enzyme called AKR1C1 binds NECTIN4 and slows down ADC uptake. 2️⃣ ADC gets in, but leaves again: instead of being routed to the lysosome, where the toxic payload is supposed to be released, the ADC gets exported out of the cell. 💡 The solution: Blocking AKR1C1 reversed both steps, restoring uptake, payload delivery, and drug sensitivity in preclinical models. Take home message: 🔹 High target expression is only half the bet, the other half is whether the ADC actually reaches the lysosome. 🔹Tumors can escape not only by reducing expression of the target, but also by disabling the following steps required for payload delivery. I am Sebastian, founder of PAIA Biotech, posting regularly about developability of biologics and technologies for assessing molecules in early discovery.

At ASCO last week, a Nectin-4 ADC posted a 42% confirmed response rate overall. In the 18 patients who had already received a topoisomerase-I ADC, it posted zero. That is Lilly's LY4052031, NEXUS-01, abstract 4508. Same drug, same dose, same protocol; the only thing separating responders from non-responders was prior exposure to that payload class. SATEEN points the same way in HER2+ breast, where sacituzumab govitecan plus trastuzumab after T-DXd produced one confirmed PR in 27 patients. Different sponsors, different constructs, the same payload-class question underneath. The ASCO 2026 ADC story is still being told largely through target novelty e.g., DLL3, SEZ6, c-MET, B7-H3, Claudin18.2, Nectin-4, and several will earn their place on tumour distribution alone. However, the question we need to ask is whether we are building increasingly differentiated antibodies around increasingly similar pharmacology. Most of the near-clinic expansion still sits on a small number of payload classes, topo-I above all, which creates a concentration risk: different antigen, different tumour, different developer, the same selection pressure underneath. This is where target-switching becomes an incomplete strategy. The resistance mechanisms, ABCG2 and P-glycoprotein efflux, topoisomerase-I alteration, enhanced DNA-damage repair, are responses to the payload, not to the antibody. The antibody finds the cell and the linker releases the drug, but the payload still has to kill, and a tumour that has adapted to a TOPO-Ii warhead does not check which antibody delivered the next one. The address changed; the package did not. Both readouts are small and heavily pretreated, so this is convergent signal, not proof of class-wide cross-resistance. Direction established, magnitude still open. However, the asymmetry points to where the next differentiation layer is moving, from antigen novelty to resistance-aware format design. Dual-payload ADCs whose aim is not more potency but less single-mechanism selection pressure. Payload chemistry built for post-topo-I sequencing rather than first-line expansion. Bispecific ADCs for avidity and antigen-low engagement, while accepting that bispecificity does nothing for payload-class resistance alone. Much of this is preclinical to Phase 1, and much is China-originated, its own signal about where format innovation now sits. For developers, an ADC portfolio built only for target diversity may still be a single-mechanism portfolio in disguise. For investors, the question is no longer whether the antigen is novel, but what resistance biology the construct effectively escapes. Map ADCs as integrated systems: antigen biology, linker processing, payload class, sequencing context and competitive exposure read together rather than separately. The next ADC winners may not be the ones with the newest targets. They may be the ones designed around the resistance mechanisms everyone else is selecting for. #ADCs #ASCO2026 #oncology

One of the cleanest mechanistic tests we’ve had in the ADC space just overturned a core assumption about how TROP2‑ and HER2‑targeting drugs actually work in metastatic breast cancer. For years, we’ve behaved as if finely slicing “TROP2‑high” and “HER2‑low” would eventually give us a reliable compass for TOP1 ADCs—better antigen quantitation, better patient selection. Mishra et al.’s PNAS study does the technically demanding version of that experiment: single‑cell, calibrated TROP2/HER2 profiling on circulating tumor cells and matched metastatic biopsies, prospectively, in patients receiving sacituzumab govitecan, datopotamab deruxtecan, or trastuzumab deruxtecan. Two key findings should make everyone working on ADCs pause: ·     Outside HER2 amplification, baseline TROP2/HER2 levels are necessary for indication, but they are not the main drivers of who gets 9–10 months of benefit versus 6–8 weeks. ·     A simple, mechanistically grounded dynamic readout—whether CTCs vanish or fall by ~80% after the first cycle—tracks durable benefit far better than any static “target‑high/target‑low” label, while ADC‑after‑ADC switching within the same TOP1 payload class rarely restores meaningful sensitivity. Scientifically, the hook is this: in these breast settings, TROP2‑ and HER2‑ADC programs are no longer target‑centric in the way we’ve been talking about them. They are payload‑centric, time‑sensitive systems with antigen‑mediated selectivity. If your development or sequencing strategy is still built on “same payload, different antigen,” Mishra et al. provide a rigorous, patient‑level argument for why that framework is now misaligned with the biology.    

Can macrophages become active partners in ADC therapy? Antibody-drug conjugates (ADCs) are designed to target antigen-positive tumor cells, yet clinical activity in antigen-low or -negative settings—like CD30⁻ DLBCL with BV or HER2-low breast cancer with T-DXd—suggests a broader paradigm. Recent evidence highlights that FcγR⁺ tumor-associated macrophages (TAMs) can internalize ADCs, process payloads intracellularly, and release them extracellularly to kill neighboring cells. This Fc-mediated macrophage route, first demonstrated by Li et al. (2017), contrasts with traditional bystander models limited to antigen-expressing tumor cells. In T-DXd, this may intersect with mechanisms such as cathepsin-mediated linker cleavage, immunogenic cell death, and STING pathway activation. Notably, FcγR-driven uptake by alveolar macrophages has also been linked to T-DXd toxicity—reinforcing their central role in payload handling. As Daiichi Sankyo explores T-DXd plus anti-SIRPα (DS-1103) in HER2-low breast cancer, one emerging hypothesis is to leverage FcγR⁺ TAMs not just for phagocytosis, but for expanding cytotoxic reach into antigen-poor regions. Furthermore, TAMs may themselves be susceptible to cytotoxic payloads—opening possibilities for reprogramming the immunosuppressive TME. This evolving model reframes ADCs not just as targeted cytotoxics, but as TME modulators—ushering in a next-generation design strategy centered on Fc tuning, payload diffusion, and macrophage biology. 📚 References: Li et al., Mol Cancer Ther 2017; Chang et al., JCI 2023; Tsao et al., Nat Commun 2025; AACR 2023 Abstract 2944

We have built an ADC target priotitization tool for K Pro. We screened 314 candidate targets for a bladder cancer program. But we're advancing just 1. Here's what happened to the other 313: 260 deprioritised by the tools’ first triage. 40 flagged as high-risk for druggability or off-target liabilities. 14 flagged as high-probability hits. 5 of those 14 killed at expert review on biological grounds. Hard conversations. 6 approved for validation. 2 currently in wet-lab validation. 1 hit ID as an ADC candidate. People ask whether AI replaces biologists in target ID. This funnel says no. The model gets us from 20,000 to 14 in days. The biologists get us from 14 to 1, and they're the reason we don't burn six months on a target that was never going to work. Posting this because most teams keep their attrition numbers private. We think the attrition is the more useful number to share.