Discussing the Toxicity Issues of ADCs

As of now, the toxicity and prevention measures of ADC in solid tumors are known as follows:

① The indications for ADCs are rapidly expanding, gradually shifting from late-stage to early-stage, and from single therapy to combination strategies.

② The toxicity characteristics of most ADCs are similar to the cytotoxicity of their payloads.

③ Certain ADCs may also exhibit unconventional and potentially life-threatening toxicities, necessitating a better understanding of these events and optimizing diagnostic and management practices.

④ The industry is seeking various pharmacological modification strategies to attempt to improve the tolerability of ADCs, including molecular changes to the antibody portion, linker, and/or cytotoxic payload.

⑤ Exploring different dosing in randomized trials and investigating dose strategies that adapt to responses can optimize the use of ADCs, maximizing their therapeutic value for each indication.

⑥ Extensive efforts are currently underway to identify toxicity biomarkers in patients receiving ADC treatment and to develop diagnostic tools that can predict and/or detect toxicity early.

01

Approved ADCs for Solid Tumors: Structure, Major Toxicities, and Interpretation of Causes

So far, the FDA and EMA have approved six ADC drugs for patients with solid tumors.The following figure describes the composition of each ADC (in terms of targeted antigens, monoclonal antibody types, payloads, and linkers), currently approved indications, and the most common toxicities observed for each drug (as shown in Figure 1 and Table 1). The adverse reaction characteristics of each ADC are typically a mix of targeted and off-target effects, with the latter often determining the maximum tolerated dose. Common adverse reactions observed to varying degrees in many ADCs include fatigue, hair loss, blood cell reduction, and gastrointestinal disorders.

Figure 1: Structures and Major Toxicities of Currently Approved ADCs for Solid Tumors

(Figure Legend: CINV, chemotherapy-induced nausea and vomiting; GGFG, Gly-Gly-Phe-Gly; DAR, drug-antibody ratio; DXd, deruxtecan; ILD, interstitial lung disease; MCC, maleimide methyl cyclohexyl-1-carboxylate; MMAE, monomethyl auristatin E; TOPO1, topoisomerase 1; Trop2, trophoblast cell surface antigen 2; VCit, valine-citrulline.)

Table 1: FDA Approved ADCs Toxicities in Patients with Solid Tumors

(Table Legend: ADC, antibody-drug conjugate; AE, adverse event; ALT, alanine aminotransferase; AST, aspartate aminotransferase; DXd, deruxtecan; FRα, folate receptor-α; GGFG, Gly-Gly-Phe-Gly; HR, hormone receptor; ILD, interstitial lung disease; MCC, maleimide methyl cyclohexyl-1-carboxylate; MMAE, monomethyl auristatin E; NSCLC, non-small cell lung cancer; TNBC, triple-negative breast cancer; Trop2, trophoblast cell surface antigen 2. a Approved only. b Includes three treatment-related deaths. c. Data available only in conference abstracts.)

Due to their large molecular weight, ADCs are usually administered via intravenous injection. In vivo experimental data indicate that this route of administration may be associated with the reduced activity of certain ADCs and severe skin toxicity. Ideally, ADCs should remain intact in circulation after injection and only release their cytotoxic payloads inside or near targeted tumor cells.

Minor changes in the structure of ADCs can often lead to significant changes in the pharmacokinetics and/or pharmacodynamics of the drug (Table 2).

After injection, ADCs typically circulate in the bloodstream as a dynamic mixture of intact ADCs (>90% of the composition), unbound drugs (or drug-linkers), and dissociated antibodies. From circulation, ADCs gradually diffuse into the interstitial space of body tissues, ultimately reaching targeted tumor cells, estimated to be about 0.1% of solid tumors..

Once ADCs reach the tumor microenvironment (TME), they are generally believed to bind to the target antigens expressed on the surface of cancer cells, undergo endocytosis, and release their payloads through chemical or enzymatic cleavage in lysosomes, ultimately leading to necrosis or apoptosis, depending on the mechanism of action of the payload and the concentration achieved at the target site.

More hydrophobic payloads (e.g., monomethyl auristatin E (MMAE) and exatecan derivatives) can diffuse beyond the target cells after dissociating from the antibody within the cell, thereby exerting a “bystander killing” effect on antigen-negative cells, which can enhance the antitumor activity of certain ADCs.

This effect also constitutes a determinant of toxicity: the released payload can also enter adjacent non-malignant cells through passive diffusion or transport protein-mediated uptake, potentially leading to off-target cytotoxicity.

In addition to the traditionally recognized mechanisms of intracellular targeted payload release in tumor cells, certain ADCs can release cytotoxic payloads without antigen involvement and endocytosis. For example, sacituzumab-govitecan has been shown to release its SN-38 payload extracellularly within the TME, a mechanism that may explain the activity and toxicity of this and other ADCs.

Table 2: Pharmacological Determinants of ADC Toxicity

As mentioned above, most components of ADCs do not reach tumor cells and gradually degrade through a combination of specific and non-specific mechanisms before elimination, including target-mediated clearance, Fcγ receptor (FcγR)-mediated uptake, and/or phagocytosis by macrophages located in various tissues.

Importantly, before degradation and excretion, the payload-linker complex released from ADCs utilizing thiol-maleimide chemistry (which accounts for the majority of approved ADCs) can react with free cysteine residues in serum albumin, resulting in long-lived albumin-linker-payload conjugates.

02

Discussing ADC Toxicity (Starting from Structure)

ADCs have a modular structure, and minor modifications to any of their key components can lead to significant changes in clinical characteristics (Figure 2). This section will analyze the relative contributions of each ADC component and the role of patient characteristics in the types and severity of observed toxicities.

Figure 2: Determinants of ADC Toxicity

(Figure Legend: Cit, citrulline; glutathione; MMAE, monomethyl auristatin E; MMAF, monomethyl auristatin F; Trop2, trophoblast cell surface antigen 2. Highly stable linkers are associated with increased incidence of certain adverse reactions, including ocular toxicity, neurotoxicity, or hepatotoxicity, depending on the specific drug.)

2.1 Payloads

According to the fundamental principles behind ADC development, the targeted antigen is expected to determine the toxicity characteristics of the drug. However, clinical experience indicates that most adverse events associated with ADCs are similar in spectrum, incidence, and severity to the payload backbone, and different ADCs sharing the same payload often exhibit similar toxicity characteristics, regardless of differences in target antigens.

These toxicities can be broadly categorized into off-target, non-tumor effects unrelated to the targeted antigen, and targeted, non-tumor effects resulting from antibody binding to homologous antigens located in non-malignant tissues.

Off-target, non-tumor toxicity dominates the toxicity characteristics of most ADCs, typically leading to adverse reaction profiles similar to that of the payload.

The key mechanisms of off-target, non-tumor toxicity for ADCs are believed to be at least partially related to the premature decoupling of the payload in systemic circulation, leading to the diffusion of free cytotoxic payloads into non-tumor compartments. This payload is typically a lipophilic molecule that can permeate cell membranes and enter non-target non-malignant cells.

As previously mentioned, some of the payloads may also bind to serum albumin and other circulating plasma proteins containing thiol groups, which can increase the half-life of the payload-linker complex and potentially lead to payload deposition in non-malignant tissues.

In addition to the mechanisms of payload dissociation from ADCs, other mechanisms are believed to mediate the exposure of non-malignant cells to cytotoxic payloads, including non-specific endocytosis, endocytosis of intact ADCs in non-malignant cells; and off-target, receptor-mediated uptake, which is caused by interactions between the Fc region of the antibody backbone and Fc receptors expressed by immune cells. The latter mechanism may be more relevant for highly stable ADCs, which are less likely to dissociate and release payloads into circulation, thus more likely to encounter intact ADCs in non-malignant tissues. Regardless of the mechanism, the extent of non-malignant cell exposure to the payload ultimately determines the tolerability of the drug, making the selection of payloads a key decision in any ADC design.

2.2 Linkers

As mentioned above, the main mechanisms leading to off-target toxicity in patients receiving ADCs may relate to the timing and localization of payload release from the conjugate. These characteristics largely depend on the stability and pharmacological structure of the linker, and therefore can have a significant impact on the toxicity characteristics of ADCs.

The ideal linker should be stable enough to deliver the payload to the intended site but also unstable enough to release an effective amount of the payload within or near the tumor.

In general, less stable linkers lead to the earlier release of free payloads into circulation, resulting in higher peak concentrations of cytotoxic agents and increased typical chemotherapy-related toxicities (such as cytopenias, hair loss, and/or gastrointestinal toxicity). More stable linkers can lead to prolonged circulation of intact ADCs and delayed release of the payload. This aspect may explain the unique toxicity characteristics of certain highly stable ADCs, some of which have been found to have limited chemotherapy-related toxicities but unexpectedly high rates of ocular toxicity. These findings suggest that a balance should be maintained when determining ADC stability, as excessive release and retention of ADC payloads may lead to unintended toxicities.

In addition to linker stability, the specific chemicals used to conjugate the payload to the antibody and the drug-antibody ratio (DAR) may also impact the toxicity characteristics of ADCs.

2.3 Antibodies

Most of the toxicity of ADCs is determined by the linker-payload complex; however, despite having the same payload and linker, ADCs can exhibit significant differences in adverse reactions due to targeted, non-tumor toxicity. These events are related to the involvement of specific targets or the accumulation of payloads in non-malignant tissues expressing ADC targets, thus varying greatly based on target antigens.

In some cases, targeted toxicity may dominate the safety profile of ADCs: An example is the EphA2-targeted MMAF ADC MEDI-547, which is associated with life-threatening bleeding and coagulation events, even at low doses, possibly related to the targeted involvement of EphA2, a receptor associated with neovascularization.

To minimize the risk of such targeted toxicities, careful selection of ADC targets is essential, prioritizing antigens expressed on tumor cells (ideally high expression) and non-malignant cells (ideally low or no expression).

Overall, although antibody-related toxicities typically do not dominate the adverse reactions of ADCs, several severe targeted, off-tumor toxicities or antibody Fc-mediated toxicities have been observed. These observations highlight the complexity of the mechanisms of action of these compounds and the correlation of each ADC component with the tolerability profile.

2.4 Patient-Related Factors

In addition to the observed differences in toxicity characteristics across various ADCs, there is also a degree of heterogeneity in the spectrum and grade of adverse reactions occurring in different patients receiving the same ADC. Multiple patient-related factors may influence the pharmacokinetics and pharmacodynamics of these drugs, including baseline organ function, the presence of comorbidities, and polymorphisms of enzymes involved in the metabolism or degradation of ADCs. Finally, ethnicity has been found to affect ADC metabolism: for example, Japanese patients have been found to have an average serum T-DXd concentration 20% higher than patients from other countries, a finding that may explain the higher incidence of ILD observed in that patient population.

03

Adverse Reactions Associated with Combination Therapy Strategies

ADCs provide multiple opportunities for combination strategies aimed at achieving additive or synergistic antitumor activity. However, these combinations also carry the risk of increased regimen toxicity, which may be due to overlapping adverse reactions or unexpected synergistic effects. In this section, we summarize the available toxicity data for ADC combinations used with different categories of anticancer agents (Tables 3 and 4).

3.1 ADC Combined with Chemotherapy

Combining different types of chemotherapy is an established method to overcome resistance and improve treatment outcomes. However, combining ADCs with chemotherapy presents certain challenges related to overlapping toxicities. Most trial data seem to indicate that the toxicity increases significantly when ADCs are combined with traditional chemotherapy, likely due to off-target effects and non-tumor effects of the ADC payload.

Table 3: Toxicity Observations of ADC Combined with Chemotherapy in Patients with Solid Tumors

(Table Legend: 5-FU, 5-fluorouracil; ADC, antibody-drug conjugate; ALT, alanine aminotransferase; AST, aspartate aminotransferase; Dato DXd, datopotamab deruxtecan; DLT, dose-limiting toxicity; ILD, interstitial lung disease; LABC, locally advanced breast cancer; MBC, metastatic breast cancer; NSCLC, non-small cell lung cancer; NR, not reported; T-DM1, trastuzumab emtansine; T-DXd, trastuzumab deruxtecan.)

3.2 ADC Combined with Endocrine Therapy

Endocrine therapy is a common treatment strategy aimed at inhibiting the growth of hormone-dependent cancers by blocking the ability of hormones to promote tumor cell growth. These agents are generally well-tolerated and are frequently administered to patients with breast or prostate cancer. Overall, combining ADCs with endocrine therapy does not seem to be associated with increased toxicity.

3.3 ADC Combined with Immunotherapy

ADCs have the potential to induce immunogenic cell death while also possessing potential immunostimulatory functions through the Fc domain of the antibody, providing a rationale for combinations with immune checkpoint inhibitors (ICIs).

So far, no signals of synergistic toxicity have been observed with ICI combinations involving T-DXd, Dato-DXd, or sacituzumab-govitecan. Further data in this field is expected from ongoing randomized phase III trials (NCT05629585, NCT05382286, and NCT05633654), which are anticipated to clarify the toxicity characteristics of combining ICIs with ADCs, excluding T-DM1.

3.4 ADC Combined with Targeted Therapy

Among currently approved ADCs, T-DM1 has the most evidence of activity and safety when used in combination with targeted agents.

T-DM1 combined with the HER2 tyrosine kinase inhibitor tucatinib was tested in a phase Ib trial and found to be tolerable, despite frequent gastrointestinal and hepatic toxicities. In the phase Ib trial of HER2CLIMB-02, T-DM1 was also tested in combination with CDK4 and CDK6 inhibitors, with no DLT observed. In a phase Ib trial involving patients with metastatic TNBC, the combination of trastuzumab-govitecan with the PARP inhibitor talazoparib resulted in multiple DLTs (most enrolled patients had febrile neutropenia). Finally, in a phase Ib trial involving patients with platinum-resistant ovarian cancer, the addition of the anti-VEGFA antibody bevacizumab to mirvetuximab-soconjugated with soravertinib produced toxicity characteristics comparable to those of ADCs alone.

Table 4: Toxicity Observations of ADC Combined with Other Therapies in Patients with Solid Tumors

(Table Legend: ADC, antibody-drug conjugate; adverse events, adverse events; ALT, alanine aminotransferase; AST, aspartate aminotransferase; ET, endocrine therapy; FRα, folate receptor-α; HR, hormone receptor; ILD, interstitial lung disease; NACT, neoadjuvant chemotherapy; T-DM1, trastuzumab emtansine; T-DXd, trastuzumab deruxtecan; TNBC, triple-negative breast cancer.)

04

Emerging Strategies to Optimize ADC Safety

Several strategies have been adopted in clinical practice to prevent or optimize the management of toxicity associated with ADCs (Figure 3).

4.1 Dose Optimization Strategies

Given the dose-dependent nature of many toxicities associated with ADCs, there is considerable interest in optimizing dosing and administration regimens to try to improve the therapeutic index. To this end, five classic dose optimization strategies have been adopted, including weight-based dosing caps, maximum treatment duration caps, dose frequency optimization, response-guided dose adaptation, and randomized dose-finding studies.

4.2 Optimizing ADC Design

Engineering ADCs and other optimization strategies are important methods to maximize the efficacy and safety of formulations. In fact, design innovations for each ADC component can fine-tune pharmacological properties and potentially impact tolerability.

4.3 Innovations in Antibody Components

Most ADCs currently approved and under investigation target antigens that are heterogeneously expressed to varying degrees in non-malignant tissues. The development of probe-drug conjugates (PDCs) provides an example of an engineering strategy that may reduce the incidence of targeted, off-tumor toxicities. In addition to masking the binding regions of antibodies, efforts are underway to silence the Fc domain of antibodies to reduce off-target, non-tumor toxicity associated with ADC uptake mediated by immune cells.

Another strategy to improve ADC delivery to tumor sites and potentially reduce the incidence of targeted toxicities is to conjugate payloads to bispecific antibodies instead of conventional antibodies. Compared to conventional antibodies, bispecific antibodies have greater selectivity and better internalization in tumor cells.

4.4 Innovations in Linker Technologies

Various strategies aimed at improving the stability of ADCs in systemic circulation are being developed to optimize safety. These methods can prepare homogeneous ADCs with predictable DARs and payload attachment sites, which may be associated with improved tolerability and more predictable pharmacokinetics. Preclinical evidence suggests that ADCs developed using these conjugation strategies retain activity and improve pharmacokinetic properties, and clinical testing is currently underway.

4.5 Innovations in Payloads

To enhance the therapeutic index of individual payload ADCs, new constructs have been developed that incorporate two different types of payloads conjugated to the same monoclonal antibody. Preclinical data have confirmed the feasibility and high antitumor activity of this approach, with constructs such as HER2-targeting antibodies conjugated to MMAE and MMAF, and FGF2-targeting antibodies conjugated to MMAE and α-amanitin.

Additionally, new payloads are being explored beyond traditional microtubule inhibitors, topoisomerase I inhibitors, and DNA intercalators. These payloads include other cytotoxic molecules (such as topoisomerase II inhibitors or agents that inhibit transcription or translation) and ADCs with unconventional payloads, such as immune-stimulating molecules, heterobifunctional protein degraders, and tyrosine kinase inhibitors.

Combination therapy of ADCs with neutralizing antibody fragments against payloads is another innovation aimed at improving ADC tolerability. Preclinical data suggest that combining HER2-targeting MMAE-based ADCs with humanized anti-MMAE Fab fragment ABC3315 in xenograft models reduced chemotherapy-related toxicities, improved weight loss in mouse models, and retained activity.

4.6 Pharmacogenomics

Identifying patients at high risk for adverse events following ADC treatment is an important step in improving the clinical management of such patients. Overall, data suggest that pharmacogenomic parameters may have relevant impacts on the safety of certain ADCs. As new ADCs are developed, incorporating pharmacogenomic profiling into the design of early trials may be a reasonable approach to ensure safety is not disproportionately affected by genetic variations in specific populations and/or individuals.

4.7 Diagnostic Tools

Another potential approach to early detection and better management of ADC-induced adverse events involves the use of wearable biosensors (WBS). The rapid development of these technologies may assist in the early diagnosis of ILD in at-risk patients, help identify acute exacerbations of ILD, and support real-time clinical decision-making.

Summary and Outlook

ADCs have improved the activity of traditional chemotherapy regimens across a range of solid tumors. Despite their ideal targeting mechanisms, most ADCs still exhibit frequent and sometimes life-threatening toxicities. Given the rapid expansion of their indications, awareness of these adverse reactions and their management, as well as efforts to prevent and mitigate ADC-related toxicities, are crucial.

These include careful attention to the risks of synergistic or overlapping toxicities when combining ADCs with other anticancer drugs, and caution in developing and testing novel ADCs in early diseases, where each indication requires a specialized balance of risks and benefits. Innovations in ADC design, pharmacogenomic testing, and WBS, along with increased attention to ADC dose optimization through dedicated prospective trials, may help realize the potential of these highly promising anticancer agents, which remain far from fully explored.

Main References

[1] Tarantino P, Ricciuti B, Pradhan SM, Tolaney SM. Optimizing the safety of antibody-drug conjugates for patients with solid tumours. Nat Rev Clin Oncol. 2023 Aug;20(8):558-576. doi: 10.1038/s41571-023-00783-w. Epub 2023 Jun 9. PMID: 37296177.

[2] Jiang, Z. et al. A multiple center, open-label, single-arm, phase II clinical trial of MRG002,

an HER2-targeted antibody-drug conjugate, in patients with HER2-low expressing advanced or metastatic breast cancer. J. Clin. Oncol. 40, 1102–1102 (2022).

[3] Yamaguchi, K. et al. Trastuzumab deruxtecan in anti–-human epidermal growth factor receptor 2 treatment-naive patients with human epidermal growth factor receptor 2-low gastric or gastroesophageal junction adenocarcinoma: exploratory cohort results in a phase II trial. J. Clin. Oncol. 41, 816–825 (2023).

[4] Nguyen, T. D., Bordeau, B. M. & Balthasar, J. P. Mechanisms of ADC toxicity and strategies

to increase ADC tolerability. Cancers 15, 713 (2023).

[5] Hasan, M. M., Laws, M., Jin, P. & Rahman, K. M. Factors influencing the choice of monoclonal antibodies for antibody–drug conjugates. Drug Discov. Today 27, 354–361 (2022).

[6] Nicolò, E. et al. Combining antibody-drug conjugates with immunotherapy in solid tumors: current landscape and future perspectives. Cancer Treat. Rev. 106, 102395(2022).

[7] Schmid, P. et al. Abstract PD11-08: PD11-08 Trastuzumab deruxtecan (T-DXd) + durvalumab (D) as first-line (1L) treatment for unresectable locally advanced/metastatic hormone receptor-negative (HR−), HER2-low breast cancer: updated results from BEGONIA, a phase 1b/2 study. Cancer Res. 83, PD11-08-PD11-08 (2023).

[8] Waks, A. G. et al. Phase Ib study of pembrolizumab in combination with trastuzumab

emtansine for metastatic HER2-positive breast cancer. J. Immunother. Cancer .