Patient, industry, and regulatory perspective on antibody‐drug conjugates dose optimization

Antibody‐drug conjugates (ADCs) represent a rapidly evolving area of oncology drug development. A critical gap remains regarding how we can widen ADC therapeutic windows that prevents patients from receiving more effective treatments for their life‐threatening illnesses. To maximize the safety and efficacy of next‐generation ADCs, it is imperative to optimize the dosages by integrating the totality of evidence, including patient perspectives via quantitative modeling. Here, we summarize the patient, industry, and regulatory perspectives on ADC dose optimization.

INTRODUCTION

Antibody‐drug conjugates (ADCs) have emerged as a rapidly evolving and promising class of oncology drugs. Currently, the US Food and Drug Administration (FDA) has approved 12 ADCs that combine the high specificity of antibodies with the high potency of cytotoxic payloads; a breakthrough in cancer therapeutics. Many ADCs have advanced into frontline regimens, yet the therapeutic window remains relatively narrow. To maximize the safety and efficacy of next‐generation ADCs, it is imperative to optimize the dosing and scheduling by integrating the totality of evidence, including patient perspectives via quantitative modeling. Here, we summarize the patient, industry, and regulatory perspectives on ADC dose optimization.

PATIENT PERSPECTIVES

The incorporation of patient perspectives and engagement has been paramount for appropriate dosage optimization. A multitude of studies have demonstrated the reliability of patient‐reported information, considering they are uniquely positioned to report on their own quality of life (QOL), symptoms, and functional status. This self‐reported information complements traditional clinician assessments.

The 2022 Lymphoma Coalition Global Patient Survey (GPS) on Lymphomas and chronic lymphatic leukemia was completed by 7113 patients from 84 countries, including patients with Hodgkin's lymphoma (HL), cutaneous lymphoma (CL), and diffuse large B‐cell lymphoma (DLBCL). ¹ These three subtypes can be treated with ADCs but have heterogeneous disease characteristics, including patient demographics, which influence the perception of optimal treatment. Among the patient groups, all ranked “cure” as the most desired treatment outcome (Table 1). Cure is most important to patients with CL, an incurable lymphoma where patients typically live a normal lifespan with periodic treatment. Interestingly, this demonstrates that patients are dissatisfied with longevity alone; they desire to be treatment‐free.

TABLE 1.

Patient data by cancer subtype: important features of a new medical treatment for lymphoma and the impact of AEs on QOL. ¹

Lymphoma type	HL	CL	DLBCL
Number of patients (n)	851	571	959
Desired treatment outcome—cure	68%	83%	77%
Desired treatment outcome—QOL	59%	61%	63%
AE impact on QOL—everyday activities	73%	43%	68%
AE impact on QOL—employment	63%	25%	52%
AE impact on QOL—social life	65%	42%	64%
AE impact on QOL—close relationships	58%	39%	59%

Abbreviations: AE, adverse event; CL, cutaneous lymphoma; HL, Hodgkin's lymphoma; DLBCL, diffuse large B‐cell; QOL, quality of life.

Patient experience of adverse events (AEs) can also influence how they prioritize QOL when choosing further treatments. Patients with CL, often treated with topical therapies with few toxicities in early‐lines, report fewer negative impacts of symptoms and side effects on QOL (Table 1). In contrast, patients with HL, a younger demographic with 84% of GPS respondents less than or equal to 50 years of age, rank cure first but not as highly as the older patients with CL and DLBCL. They place more value on improved QOL (59%) and fewer side effects (58%), likely due to aggressive first‐line chemotherapies for HL, which have good cure rates but significant short‐ and long‐term AEs. Younger patients deal with the unwanted effects of treatment long‐term, and therefore seek both survival and improved QOL.

Medically significant AEs are often different than those that patients identify as the most life‐impacting (Table 1). Toxicity persistence and QOL are more important to patients than grade. Of the three common AEs for ADCs—fatigue, peripheral neuropathy, and nausea/vomiting—all patients mentioned fatigue as having the most significant impact on wellbeing, yet 36% of patients did not receive information or support on managing their fatigue, and 58% mentioned a lack of follow‐up. Both fatigue and peripheral neuropathy can have long‐term impacts, leaving patients unsatisfied with their QOL.

Additionally, patients want to be involved in decisions related to their care (76%–78%). Patients may not be comfortable choosing a specific cancer therapy, but can express their treatment goals, personal values, and the AEs they are willing to tolerate, all of which may change over time as their treatment progresses. The foundation for optimal patient care is an effective and ongoing two‐way patient‐doctor communication that integrates patient preference into any treatment decision. Incorporating patient‐relevant outcomes into future drug development is crucial to ensure new therapies align with patient needs.

INDUSTRY PERSPECTIVE

Despite significant success for ADCs, it is worth noting the therapeutic window of currently approved ADCs remains narrow, especially when compared to other oncology drugs like targeted therapies and immunotherapies. The maximum tolerated dose of ADCs is often reached before the maximum efficacious dose, which poses a unique challenge for ADC dose optimization.

Multiple dosing strategies have been developed to minimize this issue and improve the benefit–risk profile by mitigating safety risks while maintaining efficacy. Body weight‐based dose capping (e.g., brentuximab vedotin and enfortumab vedotin) is an effective way to prevent the overdosing of heavier patients, thus minimizing the occurrence of AEs. ² If the AE of concern (e.g., nonocclusive disease) is acute and mainly driven by the maximum plasma concentration (C _max) of an ADC, optimizing the dosing schedule to a smaller and more frequent dose (e.g., gemtuzumab ozogamicin ³ and sacituzumab govitecan ⁴ ) is a viable approach to reduce risk while maintaining efficacy. However, if the AE is chronic and mainly driven by cumulative exposure (e.g., peripheral neuropathy), more frequent and lower dosing may not work; instead, capping of treatment duration (e.g., polatuzumab vedotin) can effectively mitigate this AE. ² Understanding the nature and key pharmacokinetic (PK) drivers of the AE is critical for developing an appropriate dosing approach. Response‐guided dosing is another alternative approach for personalizing the ADC dose based on patients' response, although this approach usually requires fast onset (e.g., 21 days for inotuzumab ozogamicin) of response to enable this adaptive and individualized dosing approach. Finally, randomized dose‐finding studies, especially for ADCs with relatively wide therapeutic windows that demonstrate efficacy and safety concerns across multiple doses, become increasingly important to identify an appropriate dose and schedule for late development and may increase the overall efficiency of clinical development.

Additionally, characterization of target engagement (i.e., receptor occupancy [RO]) could guide ADC dose selection, as dose escalation beyond tumor RO saturation is unlikely to provide additional antitumor activities (e.g., gemtuzumab ozogamicin). On the other hand, it is worth emphasizing that, unlike the large molecule antagonists, an optimal dose of an ADC does not always need to saturate the RO, as long as a sufficient amount of the cytotoxic payload is delivered to the tumor tissues per the mechanism of action of a given ADC.

Many of these dosing approaches are developed by the quantitative integration of preclinical pharmacology (e.g., RO), clinical PKs, pharmacodynamics (PDs), and drug efficacy and safety. A platform modeling approach for a multiscale mechanism‐based PK/PD or quantitative systems pharmacology (QSP) platform may be leveraged to integrate diverse data and knowledge of ADCs. Although the mechanism‐based PK/PD or QSP model is complex, requiring a significant amount of data, expertise, and resources, it could be extended to multiple ADCs in the portfolio once the model is established. The platform model will continue to evolve and improve with the confidence level of prediction and mechanistic insight increase over the continuous learning and confirming cycles.

Finally, a comprehensive evaluation of risk–benefit balance with seamless integration of pharmacology and disease biology is needed to maximize the therapeutic window of each ADC to determine an optimal dosing regimen. Innovative ADC dosing strategies will continue to evolve especially for “next‐generation” ADCs (e.g., novel non‐cytotoxic payloads or non‐IgG drug conjugates).

REGULATORY PERSPECTIVE

From a regulatory perspective, the investigation of a wide dose range along with the selection of multiple dose levels to be investigated in early development helps characterize the safety and efficacy of an ADC and supports exposure‐response analyses for selecting doses to be investigated in the registration trials. ⁵ , ⁶ For example, when studying the ADC trastuzumab deruxtecan, a wide dose range (0.8–8 mg/kg q3w) was studied for the escalation phase, whereas the 5.4 and 6.4 mg/kg dosages were further investigated to support exposure‐response analyses for safety and efficacy. Eventually, the 5.4 mg/kg dose was selected for the registration trial due to a favorable risk–benefit profile in the target population (HER2+ metastatic breast cancer) compared to the 6.4 mg/kg dose, which was found to be associated with similar efficacy but higher incidence of AEs. ⁷

In addition to dose level, dose schedule should be investigated in early clinical development to select a regimen that most improves the risk–benefit profile. The investigation of multiple dose levels and schedules can facilitate the identification of PK parameters that drive efficacy or safety and support the selection of a dose that improves the risk–benefit profile. At a given dose level, a fractionated regimen (i.e., smaller doses administered at a higher frequency) compared to a single cumulative dose can help delineate the effect of C _max and the associated safety profiles. For gemtuzumab ozogamicin and inotuzumab ozogamicin, the fractionated dosing regimen was found to correlate with an improved safety profile, specifically in terms of the incidence of veno‐occlusive disease, but still maintained a similar level of effectiveness. ⁸ , ⁹

Another regulatory consideration is that the unconjugated antibody may contribute to the risk–benefit profile of the ADC, which would include the unconjugated trastuzumab component of trastuzumab emtansine or trastuzumab deruxtecan. The unconjugated payload is conventionally a cytotoxic moiety where a relatively small increase in exposure may substantially increase the frequency and/or severity of AEs. In certain instances, the systemic exposure of the payload is high and may contribute directly to the safety and efficacy (i.e., SN‐38, the payload in Sacituzumab govitecan). ⁴ As such, the dose selection for an ADC should consider the pharmacology of the antibody as well as that of the payload and should be based on a thorough understanding of the PK and the PD of the antibody portion, the unconjugated payload, and the intact ADC.

The use of PD data, such as RO and target engagement, when available, should also be included in the selection of dosages for registration trials. For gemtuzumab ozogamicin, a dose of 3 mg/m² achieved maximum target occupancy with higher doses (i.e., the initially approved 9 mg/m²) not achieving any additional PD response. ⁸ With the launch of Project Optimus, the FDA expressed openness to embrace novel approaches, such as QSP models, to support dose optimization for oncology drugs. To date, QSP models have rarely been leveraged to guide regulatory decisions; however, these models have the capability to describe the complex mechanism of action of ADCs and their interactions with tumor cells and the microenvironment. These models are particularly useful in capturing tumor characteristics that dictate therapeutic response and could help with supporting dose optimization from one type of tumor to another across treatment lines or in combinations with other agents.

FUTURE DIRECTION

Although different areas of emphasis are placed on ADC dose optimization from patient, industry, and regulatory perspectives, all three parties share a common goal: identifying an appropriate dose and schedule for an ADC to maximize efficacy and mitigate unwanted AEs. With cancer gradually becoming a chronic disease, long‐term patient tolerability becomes an increasingly important factor to consider when making a dose selection decision. Additionally, medically significant AEs reported by investigators often differ from the AEs identified as most life‐impacting by patients. Toxicity persistence and QOL are continually ranked as more important to patients than the grade of disease. Therefore, it is increasingly recognized that considering a patient perspective (e.g., patient‐reported outcomes [PROs]) in dose finding studies is imperative. In 2022, Friends of Cancer Research developed a white paper highlighting key considerations for collecting PROs into dose‐finding studies. ¹⁰ Integrating patient perspectives into dosing strategies, alongside innovative dose finding study designs (e.g., randomized dose finding) and modeling and simulation approaches, will be crucial. This combined effort, involving patients, clinicians, industry and regulatory experts, promises to unlock the full potential of ADCs, ensuring treatments align with patient needs while maximizing benefit–risk ratios.

FUNDING INFORMATION

Z.Z. was supported by the National Institute Of General Medical Sciences of the National Institutes of Health under Award Number R16GM146679. All other authors received no financial support for the research, authorship, and/or publication of this article.

CONFLICT OF INTEREST STATEMENT

C.L. and M.Z.L. are employees of Genentech, Inc. and receive F. Hoffmann‐La Roche Ltd. stocks/stock options. S.H. was formally an employee of the FDA and currently an employee of Astellas Pharma US, Inc. All other authors declared no competing interests for this work.

Hamed S, Li C, Liao MZ, Warwick L, Zhou Z. Patient, industry, and regulatory perspective on antibody‐drug conjugates dose optimization. Clin Transl Sci. 2024;17:e13759. doi: 10.1111/cts.13759