ADCs or: How I Learned to Stop Worrying and Love Chemotherapy

Over 80 years ago, the dawn of medical therapy for advanced cancers was ushered in by the rapid translation of battlefield war gases into the use of nitrogen mustards to treat patients with lymphoma (1). In the next several decades, innovative chemical advances enabled curative approaches for lethal childhood leukemias and germ cell tumors, among others. However, the successful movement of these drugs into the most common malignancies plateaued, hindered by transient antitumor effects, stubborn chemoresistance mechanisms, and increasingly severe toxicities. In the era of modern drugs targeted selectively to mutated oncoproteins, the term “chemotherapy” has devolved from a “hopeful chance” to a “scourge of last resort.”

Recently, however, the field of chemotherapeutics has been revived with the development of antibody–drug conjugates (ADC) that offer the promise of increased and more targeted delivery of such “toxic” chemotherapy payloads to both increase tumor kill and spare damage to noncancerous tissues. The initial development of these compounds has focused on the use of tumor-enriched antigens such as human epidermal growth factor receptor 2 (HER2) or trophoblast cell-surface antigen 2 (TROP2) to facilitate drug delivery. These agents have shown impressive clinical success with the approval of over a dozen ADCs by the FDA as of early 2023 and hundreds more in development. The benefits seen with ADCs beg the question: Can we now exploit these effective delivery devices to bring even more powerful “chemotherapies” to bear on refractory cancers? In this issue of Cancer Discovery, Weng and colleagues (2) tackle the question of whether and how it might be possible to take an already immensely effective approach of topoisomerase I (TOP1)–targeting payloads linked to tumor antigen selective antibody and utilize a more potent but potentially more toxic payload: exatecan. Can such advances in “chemotherapy” that are made within the confines of the ADC platform yield more effective therapeutics for patients?

An ADC consists of three canonical components—a cytotoxic payload, a linker to enable its site-directed release, and a biomarker-specific mAb. Unlike traditional chemotherapy, the dose of payload delivered to tumor cells is in part limited by the bulky nature of mAbs, necessitating the use of extremely potent cytotoxic drugs even when multiple payloads are affixed per mAb. Indeed, currently approved ADCs utilize highly potent agents such as inducers of DNA cleavage (calicheamicins), microtubule inhibitors (auristatins and maytansinoids), or TOP1 inhibitors (camptothecins). Among all ADCs, those with TOP1 inhibitor–based payloads have made a tremendous clinical impact in recent years. For example, fam-trastuzumab deruxtecan (T-DXd), which uses an exatecan derivative payload, was first approved for HER2-positive breast and gastric cancers (3, 4). More recently, this drug has shown activity in non–small cell lung cancer, HER2-low breast cancer, salivary gland cancer, and urothelial cancers, among others (5, 6). Another ADC in this class is the TROP2-targeting sacituzumab govitecan (IMMU-132), which uses SN-38 (the active metabolite of irinotecan) and is approved for metastatic breast and urothelial cancers (7, 8). Despite this success, the efficacy of these ADCs in certain tumors has been modest, and the development of treatment resistance as well as observed clinical toxicity patterns remain important obstacles to overcome. Emerging data have suggested that the basis for insensitivity is not solely found in the antigen target, and payload factors have emerged as they have for conventional chemotherapy (9).

Weng and colleagues have addressed the improvement of the payload by incorporating the more potent cytotoxin exatecan in their ADC design. Exatecan was previously established to be a more potent TOP1 inhibitor than SN-38 or DXd (10, 11). It is also less susceptible to efflux by the multidrug resistance transporters of the ATP-binding cassette family, which are common mechanisms of resistance to many chemotherapies and do target DXd and SN-38. However, exatecan incorporation in ADCs was prevented by the hydrophobi­city of the drug and aggregate formation during antibody conjugation. One way of potentially overcoming this hurdle was through modifying the linker. Indeed, linker composition can strongly affect toxicity, stability, and potency and has been crucial to the overall development of ADCs by preventing promiscuous cleavage events. Weng and colleagues investigated how modifications to the maleimidocaproyl and para-aminobenzyl (pAB) spacer regions affect the aggregation of antibody–exatecan conjugates. They demonstrate that addition of polysarcosine with 10 units (pSAR10) to the pAB spacer generated low levels of aggregation and allowed for ADCs with a high drug-to-antibody ratio. Similar pSAR-based hydrophobicity masking of exatecan was also tested earlier, with promising outcomes in preclinical models (11). These data highlight how a better understanding of linker chemistry can permit the generation of antibody conjugates with safer and more potent drugs targeting a wide range of tumor types.

Weng and colleagues illustrate this concept by conjugating their modified linker and exatecan payload with multiple antibodies. They observe modest to significant benefits over existing ADCs based on trastuzumab for HER2, patritumab for HER3, sacituzumab (hRS7) or datopotamab for TROP2, and DS-6000 for cadherin 6 (CDH6). The exatecan-based ADCs show efficacy in different tumor types with high to low tumor biomarker expression. The ability to build such highly potent ADCs using the same linker and payload underscores the importance of characterizing more specific and unique biomarkers to expand the patient populations that can benefit from these drugs—an issue further emphasized by the variability in response to the exatecan ADCs. Although certain exatecan ADCs exhibited a striking response in T-DXd– or IMMU-132–resistant tumors, in other models the benefit was minimal compared with currently approved ADCs. In such cases, it becomes essential to carefully assess the long-term response, development of resistance, and toxicity for optimal selection of treatment.

A notable limitation of the current TOP1 inhibitor–based ADCs is their toxicity profile. Major adverse events in T-DXd–treated patients include interstitial lung disease/pneumonitis, and neutropenia. Neutropenia is also commonly observed with sacituzumab govitecan. Several factors can affect the toxicity profile of ADCs including the linker and payload chemistry, antibody target expression in nonmalignant tissues, and tumor type (12). In addition, there are instances of off-target toxicities seen in the clinic. Weng and colleagues have demonstrated the safety of their exatecan-based ADCs in certain animal systems, but careful analysis in early-phase clinical trials will be essential to determine the true therapeutic index of these drugs.

Paul Ehrlich, the German physician and scientist who first coined the term “chemotherapy,” later in his life called for the need to “aim chemically.” The ADC platform represents a major leap toward achieving this vision and with it, renewed interest in the power of chemotherapeutics. Specific advancements in linker and payload chemistry are moving the field of “targeted chemotherapy” toward an era of fourth-generation ADCs. With these improved methods, other payload classes can also be explored in the clinic soon. For instance, radionuclides (13) may be used as payloads in tumors that have developed resistance to standard cytotoxic drugs. Learning from the past, we do expect to encounter cancers that are able to evade even the most sophisticated antineoplastic agents. However, incorporating rational hypotheses about treatment resistance mechanisms can inform the design of better drugs that hold the promise of broadly improving outcomes for patients with cancer. With regard to the modern use of chemotherapy, sometimes tilling old soil can yield new fruit.

J.Z. Drago reports grants from AstraZeneca and personal fees from Virology Education, Intellisphere LLC, Biotheranostics, Genagon, AmMax Bio, and NuProbe outside the submitted work. S. Chandarlapaty reports grants from Daiichi Sankyo, Ambrx, and Paige.ai, other support from Odyssey Biosciences and Totus Medicines, personal fees from Novartis and Lilly, and grants and personal fees from AstraZeneca outside the submitted work. No disclosures were reported by the other author.