Abstract and Introduction
Abstract
Antibody-drug conjugates (ADCs) are one of the fastest-growing oncology therapeutics, merging the cytotoxic effect of conjugated payload with the high specific ability and selectivity of monoclonal antibody targeted on a specific cancer cell membrane antigen. The main targets for ADC development are antigens commonly expressed by lung cancer cells, but not in normal tissues. They include human epidermal growth factor receptor 2, human epidermal growth factor receptor 3, trophoblast cell surface antigen 2, c-MET, carcinoembryonic antigen–related cell adhesion molecule 5, and B7-H3, each with one or more specific ADCs that showed encouraging results in the lung cancer field, more in non–small-cell lung cancer than in small-cell lung cancer histology. To date, multiple ADCs are under evaluation, alone or in combination with different molecules (eg, chemotherapy agents or immune checkpoint inhibitors), and the optimal strategy for selecting patients who may benefit from the treatment is evolving, including an improvement of biomarker understanding, involving markers of resistance or response to the payload, besides the antibody target. In this review, we discuss the available evidence and future perspectives on ADCs for lung cancer treatment, including a comprehensive discussion on structure-based drug design, mechanism of action, and resistance concepts. Data were summarized by specific target antigen, biology, efficacy, and safety, differing among ADCs according to the ADC payload and their pharmacokinetics and pharmacodynamics properties.
Introduction
Antibody-drug conjugates (ADCs) are novel therapeutic agents composed of a monoclonal antibody (mAb) carrying a cytotoxic drug (payload) through a linker. This growing class of anticancer drugs has been developed to improve the therapeutic window of conventional chemotherapy through selective delivery to tumor cells expressing the mAb target antigen, thereby limiting the potential off-target systemic toxicities.[1,2] Currently, the evaluated ADCs carry highly potent cytotoxic payloads, which would not be suitable for systemic use because of high systemic toxicity and a low therapeutic index. Subsequently, the payload's short half-life is desirable to prevent potential off-target toxicities. The linkers have to be highly stable in plasma to prevent the payload from being released in a nonspecific manner while circulating in blood,[3,4] while they must enable efficient payload release at tumor sites. For each ADC, the drug-to-antibody ratio (DAR), corresponding to the median number of payload moieties linked to each mAb (ranging from 2 to 8 among the ADCs with current approval) is expected to reflect drug potency and higher cytotoxicity[5] (Table 1). However, although ADCs with higher DAR have more potent in vitro activity, in vivo models have shown faster hepatic plasma clearance,[6,7] possibly related to the overall size of the compounds making them more susceptible to clearance by the liver.[7] Consequently, increased hepatic clearance might lead to a lower tumor exposure to the drug as well as a potentially lower therapeutic index.
Mechanism of Action of ADCs
The first requirement for ADC activity is adequate tumor penetration. Indeed, ADCs are able to reach target cancer cells via passive diffusion after extravasation, but mAbs are large molecules (approximately 150 kDa) and may have limited ability to passively diffuse across the tumor vasculature, particularly in tumors with high interstitial pressure.[8,9] This mechanism implies that only a small fraction of the administered ADC dose actually binds tumor cells, usually with a pharmacologic peak in drug concentration at tumor sites 1–2 days after dosing in mouse models.[10–12]
The second crucial step consists of target engagement, namely the ADC binding to the target antigen on the cell surface through the specific mAb.[13] This step might also result in a specific antibody-mediated immune activity, by blocking ligand binding, interfering with target protein dimerization, favoring target internalization and degradation, or through Fc-mediated antibody-dependent cellular cytotoxicity (ADCC).[14–16] In addition, for some targets that are upstream of oncogenic signaling pathways (eg, human epidermal growth factor receptor 2 [HER2]), the antibody may exploit an intrinsic activity, notably via modulation of downstream signal transduction.[17,18] The use of Ab targets that harbor an oncogenic function favors internalization and ubiquitination of target-ADC complex[17–19] and results in a lower probability of downregulation of the target expression as a resistance mechanism.
After target engagement, the ADC-antigen complex is internalized via antigen-dependent endocytosis and trafficking along endosomes and/or lysosomes.[20,21] The payload release occurs within early endosomes (those formed immediately after the invagination of the plasma membrane and characterized by low pH), for those with acid-cleavable linkers, and in late endosomes (formed from the maturation of early endosomes, characterized by a higher pH) or lysosomes for those with protease-cleavable linkers as well as for noncleavable linkers after proteolytic degradation.[4,22] In addition, cleavable linkers can release their payloads extracellularly within the tumor tissue, because of the presence of a redox tumor microenvironment (TME), with low pH and high concentration of proteases.[3,23] Together with the payload passive diffusion across the cell membrane after internalization and processing, payload release in the TME is responsible for the so-called bystander effect, a cytotoxic effect exerted on neighboring cells independent of the target antigen expression.[24–26] However, not all ADCs exhibit the bystander effect. A higher bystander effect will be supported by a higher DAR and for ADCs with cleavable linkers, while the properties of the TME will regulate the cleavage of the linker and payload release.[27] Finally, the tumor cell has to be sensitive to the cytotoxic payload within the ADC.
Leading Targets for ADCs in Lung Cancer
HER2. HER2 is a receptor tyrosine kinase (RTK), considered as a therapeutic target in non–small-cell lung cancer (NSCLC). Indeed, alterations of the corresponding oncogene ERBB2 may result in gene amplification (defined as HER2/CEP7 ratio ≥2, occurring in approximately 2%-5% of cases), gene mutations (2%-4%), or protein overexpression.[28] HER2 mutations encompass heterogeneous alterations distributed in the extracellular, tyrosine kinase, or transmembrane domains. The dominant form of all mutations is an exon 20 insertion—and the most common variant is a 12-base-pair encoding YVMA in a frame within the kinase domain. As HER2 has no specific ligand, downstream signaling occurs through the formation of homodimers and heterodimers with HER1 and human epidermal growth factor receptor 3 (HER3).[29,30] Since HER2 is less subject to internalization and degradation than other HER family members that demonstrate a higher receptor turnover, it can remain activated for a long time on the cell membrane and represents a theoretically ideal target for ADC development.[29]
Trastuzumab emtansine (T-DM1) is an ADC conjugating the anti-HER2 mAb trastuzumab with the microtubule inhibitor maytansine derivative emtansine (DM1) through a noncleavable thioether linker (Figure 1; DAR 3.1).[31] In a phase II basket trial, 18 patients with HER2-mutant advanced NSCLC—either treatment-naive or chemotherapy pretreated—received T-DM1 with an overall response rate (ORR) of 44% and median progression-free survival (mPFS) of 5 months. No predictive value of HER2 immunohistochemistry (IHC) expression was observed in this trial.[32] Updated results, with additional 31 patients enrolled (HER2-mutant or -amplified), showed an ORR of 51%, with an mPFS of 5 months.[19] Another phase II study evaluated T-DM1 specifically in pretreated HER2-overexpressing NSCLC, with limited responses (ORR 20%), confined to the IHC3+ cohort (n = 20).[33] After the evidence of epidermal growth factor receptor (EGFR) upregulation in tumor cells upon T-DM1 treatment, a phase II trial (TRAEMOS) was conducted combining T-DM1 and osimertinib in EGFR-mutant patients with HER2 overexpression and/or amplification after progression on osimertinib (n = 27), but was stopped early because of very limited efficacy (ORR 13%, mPFS 2.8 months) demonstrated at the preplanned interim analysis.[34]
Figure 1.
Overview of ADCs for lung cancer. In the illustration are reported the main ADCs for lung cancer treatment. For each ADC molecule, the corresponding payload and cell surface antigen are represented. ADC, antibody-drug conjugate; CEACAM5, carcinoembryonic antigenrelated cell adhesion molecule 5; HER, human epidermal growth factor receptor; TROP2, trophoblast cell surface antigen 2.
Trastuzumab deruxtecan is another HER2-directed ADC, where the mAb is conjugated through a cleavable linker—with a high DAR of 8—of the exatecan derivative topoisomerase I inhibitor deruxtecan (DXd; Figure 1).[35] T-DXd received accelerated approval by the US Food and Drug Administration (FDA) for the treatment of HER2-mutant NSCLC who have received a prior systemic therapy.[36,37] The approval was based on the results of the phase II DESTINY-Lung01 trial, which included two cohorts of patients: HER2-overexpressing and HER2-mutant advanced NSCLC. Ninety-one patients in the HER2-mutant cohort received T-DXd 6.4 mg/kg once every 3 weeks, with an ORR of 55%, a disease control rate (DCR) of 92%, an mPFS of 8.2 months (95% CI, 6.0 to 11.9), and a median overall survival (mOS) of 17.8 months.[38,39] Responses were observed across different HER2 mutation subtypes, as well as across HER2 expression levels or regardless of the presence of HER2 amplification. One potential reason for efficacy in HER2-mutant NSCLC regardless of expression is the enhanced HER2 internalization in the presence of a mutation.[19] Grade ≥3 drug-related adverse events occurred in 46% of patients, the most common being neutropenia and anemia. Of note, drug-related interstitial lung disease (ILD) occurred in 26% of patients.[38] The phase II DESTINY-Lung02 trial randomly assigned patients with HER2-mutant NSCLC to receive T-DXd 5.4 or 6.4 mg/kg once every 3 weeks. The results were consistent between the two cohorts, with a more favorable safety profile with 5.4 compared to 6.4 mg/kg (drug-related ILD in 5.9% and 14.0%, respectively).[40] Clinical trials are ongoing to evaluate the use of 5.4 mg/kg of T-DXd in the frontline setting (DESTINY-Lung04), as well as in combination with PD-L1 inhibitors (Table 2).
HER3. HER3 is another member of the RTKs HER family, commonly expressed (in up to 83%) in NSCLC.[29,41] The EGFR:HER3 heterodimerization has been reported as a mechanism of resistance to EGFR tyrosine kinase inhibitors (TKIs) by potentiating ERBB3-dependent activation of oncogenic signaling.[42]
Patritumab deruxtecan is an HER3-targeted ADC, composed of a fully human anti-HER3 IgG1 mAb conjugated with the exatecan derivative DXd through a cleavable tetrapeptide-based linker (Figure 1; DAR 8).[43] In the phase I dose expansion study HERTHENA-Lung01, 57 patients with EGFR-mutant adenocarcinoma progressing on previous EGFR TKIs and one or more prior platinum-based chemotherapy regimens in cohort 1 received patritumab deruxtecan 5.6 mg/kg once every 3 weeks.[41] Confirmed ORR was 39%, and responses were observed across resistance mechanisms. HER3 membrane expression was observed in all patients, and ORR was observed across a wide range of baseline tumor HER3 membrane H-scores. DCR was 72% and mPFS was 8.2 months (95% CI, 4.4 to 8.3). Grade ≥3 treatment-related were observed in 54% of patients. Treatment-related ILDs occurred in 7% of patients treated at the recommended dose for expansion.[41]
In the dose expansion cohort 2, 47 patients with pretreated unselected NSCLC (without EGFR activating mutations) were included. ORR was 28% and mPFS was 5.4 months (95% CI, 3.9 to 12.7).[44] Patritumab-DXd has received breakthrough designation from the FDA.
Trophoblast Cell Surface Antigen 2. Trophoblast cell surface antigen 2 (TROP2) is a transmembrane glycoprotein of the epithelial cell adhesion molecule (EpCAM) family, frequently expressed in many tumors but sporadically in normal tissue.[45]
Datopotamab deruxtecan (Dapo-DXd) is an ADC composed of an anti-TROP2 mAb and the DXd payload with a tetrapeptide-based cleavable linker (Figure 1; DAR 4).[46] The first-in-human TROPION-PanTumor01 study included 175 patients with previously treated NSCLC, unselected for TROP2 expression, in the dose expansion cohort to receive Dato-DXd at the dose of 4, 6, or 8 mg/kg once every 3 weeks. ORR was 23%, 21%, and 25%, and mPFS was 4.3, 8.2, and 5.4 months, respectively.[47–49] Of note, activity was observed in the subset analysis of patients with actionable genomic alterations (AGA, n = 34, 85% with EGFR mutations), showing an ORR of 35% and a medium duration of response (mDOR) 9.5 months.[50] TEAEs in the overall study population were mainly dose-dependent, with ILD emerging in 15% of patients at 8 mg/kg dose compared with 2% at 4 mg/kg. In actionable genomic alteration population, drug-related grade ≥3 TEAEs were observed in 38% of patients, and only one case of adjudicated ILD (grade 5) occurred in a patient who received Dato-DXd 8 mg/kg.[49]
The 6 mg/kg dose was then selected for the ongoing registrational phase III randomized trial TROPION-Lung01, evaluating Dato-DXd versus docetaxel in previously treated NSCLC with or without genomic alterations.[51] Combinatorial phase Ib studies are ongoing to evaluate safety and efficacy of Dato-DXd with PD-L1 inhibitors.[52] Initial results from the TROPION-Lung02, evaluating the combination of Dato-DXd with pembrolizumab alone (doublet) or with platinum chemotherapy (triplet), have been presented.[53] Notably, the combinatorial arm was designed with only carboplatin or cisplatin, assuming the ADC as replacement of the platinum-doublet compound. Overall, safety results were manageable, with stomatitis and nausea being the most frequent TEAEs, mostly grade 1–2. In the first-line setting, ORRs were 62% and 35% in the doublet and triplet cohort, respectively.[53] Of note, patients in the trials with Dato-DXd are not selected by TROP2 expression on tumor tissue, although further retrospective evaluations are planned to evaluate whether expression correlates with clinical efficacy.
Sacituzumab govitecan is another TROP2 targeted ADC, composed of an anti-TROP2 mAb, the topoisomerase I inhibitor, irinotecan active metabolite camptothecin derivative, SN-38, and a cleavable linker (Figure 1; DAR 7.6).[54]
The phase I/II IMMU-132-01 trial included overall 495 patients with refractory metastatic cancers to receive sacituzumab govitecan at a dose of 8, 10, or 12 mg/kg on day 1 and 8, of 3 weeks cycle. Among 54 patients with NSCLC, ORR was 16.7%, mPFS was 4.4 months, and mOS was 7.3 months.[55,56] Of note, more than 90% of tumor specimens evaluable in this study had high TROP2 IHC expression. Sacituzumab govitecan is currently under evaluation versus docetaxel in a phase III trial in pretreated NSCLC, and in combinatorial approaches including with immune checkpoint inhibitors and PARP inhibitors (ClinicalTrials.gov identifier: NCT04826341). In the same IMMU-132-01 study, ORR was 17.7%, mPFS 3.7 months, and mOS 7.1 months in patients with small-cell lung cancer (SCLC; n = 62).[56,57]
MET. The transmembrane RTK c-MET, activated upon binding of hepatocyte growth factor, is dysregulated in several cancer types. In NSCLC, MET overexpression is observed in 30%-50% of cases, whereas MET amplification (1.5%) or MET exon 14 skipping mutations (3%) are identified as primary driver alterations—these molecular features not being found to be associated.[58–60] In addition, MET amplification can occur as a resistance mechanism in EGFR-mutant NSCLC resistant to EGFR TKIs. In the available literature, variable molecular definitions of MET amplification have been adopted, affecting consistency and comparability across trials.[61]
Telisotuzumab vedotin (Teliso-V) is an ADC composed of the humanized ABT-700 mAb targeting c-MET linked to the microtubule inhibitor monomethyl auristatin E (MMAE) via cleavable valine-citrulline linker (Figure 1; DAR 3.1).[62]
The first-in-human trial conducted with Teliso-V at 2.4–3 mg/kg once every 3 weeks, included 16 patients with NSCLC and MET overexpression. ORR was 19%, and mPFS was 5.7 months, with grade ≥3 TEAEs in 17% of patients.[63] In the phase II trial Lung-MAP Sub-study S1400K, including 23 patients with MET-positive squamous NSCLC treated at the recommended dose of 2.7 mg/kg once every 3 weeks, the ORR was 9%.[64]
In a phase Ib study, Teliso-V was evaluated in combination with erlotinib in patients with c-MET-positive NSCLC.[65] In the population of EGFR mutation (del19 or L858R)-positive patients at resistance to EGFR TKI (n = 28), ORR was 32.1% and DCR was 85.7%, with an mPFS of 5.9 months. In those EGFR mutation-positive patients with high c-MET expression (n = 15), ORR was 52.6%. ORR (39%) and DCR (100%) were higher in patients who did not receive a previous third-generation EGFR TKI compared with those who received previous osimertinib (ORR 27% and DCR 73.3%). Conversely, ORR was 40% and DCR was 80% in EGFR WT patients (n = 5).[65]
The Luminosity trial was conducted to evaluate c-MET potential predictive biomarkers for Teliso-V treatment. In the nonsquamous EGFR WT subgroup, the ORR was 53.8% in c-MET high and 25% in c-MET intermediate (c-Met overexpression high: ≥50% tumor cells at 3+ intensity; intermediate: 25%-49%). This cohort is currently being expanded into stage 2 enrollment.[66,67] The results were less consistent in the cohorts of nonsquamous EGFR-mutant (ORR 13.3%) and squamous (ORR 14.3%).
The combination of Teliso-V and osimertinib has been evaluated in 25 NSCLC patients with EGFR-mutant tumors who had developed resistance to osimertinib and had c-Met overexpression high/intermediate, with ORR of 58% and grade ≥3 AEs 20%.[68]
Other Targets in Development for ADCs in Lung Cancer
Different ADCs targeting other transmembrane proteins or membrane receptors, including carcinoembryonic antigen–related cell adhesion molecule 5 (CEACAM5), B7H3, and delta-like protein 3 (DLL3), are currently in development in specific clinical trials in lung cancer (Table 2). Extensive discussion on these new targets is reported in Appendix 1 (online only).
Focus on ADC Toxicities
The use of ADCs in the treatment of lung cancer has been challenged by their potential to cause toxicities. Although modern ADCs have improved safety profiles, they can still cause debilitating and potentially fatal adverse events, such as pulmonary, hepatic, neurologic, and ophthalmic events. These toxicities are mainly due to off-target effects resulting from the premature release of the ADC payload in circulation or TME (payload-related) or the binding of the ADC to noncancerous cells expressing the target antigen (target-related).[69,70]
Hence, the toxicity profile of ADCs can be influenced by the expression pattern of the target antigen, and toxicities can occur either on target or off target. Meta-analyses have shown that specific payloads, such as MMAE, are associated with anemia, neutropenia, and peripheral neuropathy, while others, such as DM1, are associated with thrombocytopenia and hepatotoxicity, regardless of the target antigen.[70] In addition, T-Dxd has been related to high rate of ILD incidence across various types of tumors, while auristatin-based ADCs have been linked to high-grade neurologic and ocular toxicities (MMAF and also DM4).[71]
Toxicity profiles can also vary among different ADCs, even with similar payloads and linkers. For example, T-Dxd, patritumab-DXd, and Dato-DXd have different toxicity profiles despite having the same payload and linker. In particular, the incidence of ILD/pneumonitis in patients with lung cancer has been reported at variable rates across ADC clinical trials.[72] The pathogenesis of ADC-related ILD and pneumonitis is not fully understood, but multiple and potentially overlapping mechanisms are proposed:[73,74] (1) bystander effect: a high affinity of the target antibody (HER2, TROP2, and HER3) for tumor cells might lead to increased exposure of normal lung tissue to the payload (through bystander effect), leading to a higher risk of dose-dependent and direct toxicity; (2) local inflammation: the antibody might bind to cancer cells and normal cells (expressing the target antigen) in the lung, leading to direct damage; (3) ADCC: through cell binding, the ADC can activate immune mediators leading to inflammatory processes and finally development of ILD.[73]
Since ILD can be a life-threatening adverse event, adequate patient selection, as well as early and accurate diagnosis and treatment, is crucial in the real-world setting. The discontinuation of the ADC may be necessary in symptomatic ILD, and reintroduction may be considered in asymptomatic cases at modulated doses after complete resolution of ILD. Corticosteroids remain the mainstay of ILD treatment, and the dose depends on the severity of the event.[75]
Strategies to Improve ADC Efficacy
Biomarker Selection. The activity of an ADC should theoretically depend on the effective presence of the target on tumor cell surface. However, treatment strategies using ADCs in lung cancer pursue either biomarker-driven or biomarker-agnostic approaches.
The ADC strategy in HER2- and MET-driven tumors requires the identification of enrichment in target expression on tumor tissue. However, accurate biomarker selection is still an open issue. Teliso-V is being evaluated in MET-overexpressing tumors, independently of MET amplification. However, different IHC levels and cutoffs might affect differential efficacy results across different trials, as well as the reproducibility in future clinical practice.[61]
Conversely, only the presence of HER2 mutations and not overexpression has been associated with response to anti-HER2 ADCs specifically in NSCLC. This is far different from the results observed in breast cancer, where HER2-low or -negative tumors showed responses to anti-HER2 ADCs currently in development.[76] So far, such differential results remain unexplained, but are felt to be due to different tumor biology and notably cancer cell dependency on the specific oncogenic signalling—suggesting that stronger and established drivers might represent better targets.[39]
A biomarker-agnostic approach is instead applied today for other ADCs that bind to targets characterized by a high expression prevalence (ie, HER3, TROP2, B7H3, and nectin4) in lung cancer cells. To our knowledge, to date, no benefit for target enrichment has been shown for these compounds in exploratory analyses of early clinical trials. Although TROP2 enrichment was predictive of tumor inhibition with Dato-Dxd in in vitro models,[46] no significant differences in survival were observed according to TROP2-high, -medium, and -low H-scores (evaluating intensity and percentage of cells staining positive) in clinical trials.[77] These findings might suggest lack of selectivity of the TROP2 diagnostic antibodies used, and a refinement of TROP2 quantification could help to overcome this potential assessment bottleneck. Additional factors impairing the validity of current TROP2-related enrichment strategies include the expected dynamicity of receptor turnaround, and heterogeneity of target expression, which cannot be completely assessed with a small tumor biopsy specimen.
On the basis of these considerations, treatment optimization would benefit from extended assessment on tumor biopsies, immediately before treatment selection, which might encompass staining of multiple ADC targets—in a multiplex or sequential fashion—to guide ADC selection and potentially the most appropriate ADC-based strategy (eg, monotherapy, combinations, and sequencing) for each patient.
ADC Structure. There is no definitive hierarchy of key attributes that make a good ADC, as the design and selection of an ADC will depend on various factors, including the type of cancer, the target antigen, and the desired therapeutic effect.
As previously discussed, target selection is a key feature for a good ADC. Indeed, the identification of a well-established target is required to help ensuring that the ADC will effectively bind to and eliminate cancer cells.
Appropriate antibody selection is also a key attribute to consider, not only for the antibody immune-modulating properties of the antibody itself, but also for its pharmacochemical properties when linked to the payload. As the pharmacokinetic of the antibody-payload complex may largely differ with respect to the compounds alone and affect the distribution, selectivity, and bioavailability, different antibodies and payloads are tested to identify the optimal combination for a specific target.
Antibody-drug conjugation is another aspect to consider. Indeed, in nonspecific, conventional conjugation, cytotoxin binding to lysines or cysteines, which are abundant in antibody structure, can result in more off-target side effects. By contrast, site-specific conjugation is built on engineered specific sites, therefore increasing the therapeutic index.
Linker properties are another point to evaluate to build a good ADC. The presence of a stable linker is required to ensure that the payload is only released in the target cells and not in normal cells, and a specific cleavage of the linker is important to ensure that the payload is effectively released and can exert its therapeutic effect in the TME.[16]
The choice of cleavage mechanism will affect activity and safety profile of ADC. The cleavage of the linker can affect the pharmacokinetic of the payload by altering the rate and extent of release from the ADC.[78] For example, a more efficient cleavage mechanism can result in a more rapid and complete release of the payload, which can increase the therapeutic effect but also increase the risk of toxicity. Cleavage by tumor-enriched enzymes, such as cathepsin B for T-DXd, can help to increase the specificity of the ADC for cancer cells.[79]
Beyond the efforts to optimize the selection of the most appropriate mAb-cytotoxic payload-linker composition for each ADCs, research is moving toward novel approaches and payload selection. These include the use of bispecific antibodies for enhanced internalization and/or improved tumor specificity,[80] the use of small-molecule drug conjugates—instead of the antibody backbone-to increase tumor tissue penetration including CNS,[81] or the use of alternative payloads.[82] ADCs that contain a microtubule inhibitor as the cytotoxic payload (such as MMAE) have been associated with treatment-related neuropathy.[83] As such alternative cytotoxic payloads, including topoisomerase I inhibitors, may be more desirable to avoid such side effect. Another example is the pyrrolobenzodiazepine payload coupled with monoclonal antibody DLL3, which has shown unacceptable toxicity despite its high potency.[84] The development of a variety of payloads can provide a range of options for ADCs with different mechanism of action, potency, and toxicity profiles. Moreover, noncytotoxic payloads offer potential advantages over traditional cytotoxic ADCs, including reduced toxicity to healthy cells and increased specificity in targeting cancer cells. Some examples of noncytotoxic payloads in ADCs include
Radioisotopes, therefore with ADCs delivering radiation directly to cancer cells;[85]
Enzymes: ADCs carrying enzymes can be used to target and degrade specific proteins in the cancer cells, leading to cell death;[86]
Small molecules, which can be used to inhibit specific targets in the cancer cells, leading to cell death or altering cancer cell behaviour;[87]
Peptides: ADCs with peptides as payloads can be used to target and activate immune cells, leading to an immune response against cancer cells.[88]
However, these strategies are still in early stages of development and require further research and clinical testing.
Finally, DAR optimization is an important aspect of ADC development. A higher DAR typically results in a higher dose of the therapeutic payload delivered to the target cancer cells, leading to increased efficacy. However, it can also result in increased toxicity to normal cells, which can limit the therapeutic window. Therefore, it is important to carefully evaluate the DAR to optimize the balance between efficacy and safety for each ADC. This can be achieved through various methods, such as adjusting the number of drugs attached to each antibody, modifying the linker technology, or selecting a different payload.[89]
Combination Strategies. The different results obtained with T-DM1 and T-DXd, two ADCs sharing same mAb and epitope recognition, suggest that improved efficacy can be obtained by using different payloads. The sequential use of ADCs targeting the same receptor could possibly overcome payload-related toxicities. As demonstrated in HER2-positive breast cancer, the sequential use of ADCs with the same target but different payload might be a strategy to overcome resistance,[37] especially if the resistance is not target-dependent.
Besides sequential approach, the combination of ADCs with different payloads and/or with different coexisting targets could be evaluated as a novel polychemotherapy strategy to improve treatment efficacy, despite requiring particular attention on potential overlapping toxicities.
In addition, combinatorial strategies are under investigation to assess the safety and efficacy of ADCs delivered in association with partner drugs that can modulate the target antigen dynamics (eg, the combination with TKIs directed against the same ADC target to increase internalization, or other pathway regulators to upregulate target expression),[19,43] drugs that can modify the TME composition and ADC penetration (eg, antiangiogenetics),[90] immunotherapeutic agents for potential synergistic effect in immunologically cold tumors,[91] or even systemic cytotoxic agents with alternative mechanism of action.[92] Of note, the combination of ADCs with prior TKIs could also be evaluated as a strategy to overcome TKI resistance not mediated by on-target resistance.[93]
With regards to the combination with ICIs, ADCs work by triggering various processes in cancer and immune cells, including immunogenic cell death, antibody-dependent cell-mediated cytotoxicity, and dendritic cell activation, which can complement immunotherapy.[91] The synergistic potential of an ADC with ICIs can depend on factors such as the antibody used (IgG1 mediating higher ADCC than other IgG subtypes), the target antigen, the linker, and the payload (and subsequent tumor cell killing through bystander effect).[94] In a future perspective, ADC results in the advanced setting are proof of concept for transferring drug evaluation in the early disease. As such, clinical trials are being envisaged in the adjuvant/neoadjuvant setting, where ADCs, alone or in combination with ICIs, are envisaged to replace conventional chemotherapy.
Principles of ADC Resistance
On the basis of the complex and dynamic mechanism of action of ADCs, resistance to ADCs might be attributed to multiple factors, that can also co-occur in vivo (Figure 2).
Figure 2.
Proposed mechanisms of ADC resistance. Theoretical mechanisms of acquired ADC resistance are conceptualized through the three major steps of ADC mechanism of action: (1) target binding can be limited by downregulation or increased degradation of the target antigen after chronic exposure to a target-directed ADC, with subsequently less ADC-binding and internalization in tumor cells; (2) decreased endosomal/lysosomal acidification and proteolytic activity can alter intracellular trafficking of the mAb-payload complex reducing the payload intracellular release; and (3) payload activity can be impaired by direct alteration on the payload target (eg, microtubule or DNA repair mechanisms) or upregulation of drug efflux pumps, in particular for those payloads that are ABC transporter substrates, with subsequent payload ejection. Despite some of these mechanisms have been demonstrated in vitro, most of these are tough to measure or evaluate from an actual tumor biopsy as would require multiple assays to test for hard-wired genetic changes affecting any steps of ADC mechanism of action, as well as assays to evaluate whether diagnostic and therapeutic antibodies bind to the same site on a target antigen. ADC, antibody-drug conjugate; mAb, monoclonal antibody.
Primary lack of activity might be attributed to pharmacodynamic properties, primarily related to inadequate tumor penetration of the mAb-payload complex, as well as to the heterogeneity of target expression, especially in those ADCs with noncleavable linkers that have limited bystander effect activity.[95]
Acquired resistance mechanisms are instead related to the dynamic of the tumor cells and their microenvironment and might be categorized according to the three major steps of ADC functioning: target binding, intracellular trafficking of the mAb-payload complex, and payload activity.[96]
Indeed, chronic exposure to a target-directed ADC leads to downregulation of the target expression, with subsequently less ADC-binding and internalization in tumor cells.[97] The effectiveness of an ADC can depend on whether the target antigen has undergone any mutations that affect the binding site for the antibody (epitope), or if the internalization motif has changed, which can affect the rate at which the ADC is taken up into the cell. If either of these components is altered, the ADC may not bind effectively or internalize efficiently, leading to decreased efficacy.
This aspect might have less impact in ADCs with pH-dependent cleavable linkers, that retain the possibility for the payload to be released in the tumor stroma, according to the hypoxic and acidic conditions of the TME.[98]
In parallel, decreased endosomal/lysosomal acidification and proteolytic activity can occur, reducing the payload intracellular release.[99] In addition, upregulation of drug efflux pumps has been observed, in particular for those payloads that are ABC transporter substrates, with subsequent payload ejection.[100]
In conclusion, the use of ADCs for lung cancer treatment has demonstrated encouraging results. Several target antigens (HER2, HER3, TROP2, c-MET, CEACAM5, and B7-H3) have been evaluated with their specific ADCs, and some of these agents (eg, anti-HER2) are already available in clinical practice. Promising results have been obtained not only in NSCLC, but also in SCLC. Other target antigens of interest are under investigation with novel ADCs currently in early-phase clinical trials.
Appropriate patient selection and drug development approaches will help improve toxicity profiles, as well as strategies to overcome acquired ADC resistance.
J Clin Oncol. 2023;41(21):3747-3761. © 2023 American Society of Clinical Oncology
