When anticancer drugs are developed, the first-time-in-human study is initiated with dose escalation to find the best dose for the subsequent development. In this stage, the assumption of the optimal dose is the maximum tolerated dose (MTD) which was well applied to cytotoxic agents since those drugs show a steep dose-response relationship. The more drug administered, the greater the tumor cell die as far as the human body is tolerated. These days, most new anticancer drugs are targeted agents that inhibit molecular pathways of proliferation in cancer cells or inhibit their death. For these agents, dosing at the MTD is often inappropriate. Higher doses lead to off-target effects: toxicity, dose interruptions, and reduced compliance, while much lower doses result in good tumor response with much lower toxicity and better drug compliance. Nevertheless, most new targeted anticancer agents are still tested in early phase clinical trials to determine the MTD without incremental benefit and that dose is carried forward into late-stage studies. Therefore, some groups of oncologists have suggested the optimal dose of a new anticancer drug would be determined best through a randomized dose-ranging phase II trial [1]. Recently FDA and many stakeholders including Friends of Cancer Research have started to advocate the concept and implementation in the early drug development process [2, 3]. In this commentary, issues from conventional dose findings in early phase oncology trials and suggestions and recommendations by Friends of Cancer Research and FDA initiatives with a case of sotorasib are addressed.

Friends of Cancer Research (Friends) is an organization to drive collaboration among partners from every healthcare sector to power advances in science, policy, and regulation that speed life-saving treatments to patients. Friends is working to accelerate policy change, support groundbreaking science, and deliver new therapies to patients quickly and safely [4].

The Friends white paper in 2021 addressing “Optimizing Dosing in Oncology Drug Development” offered these suggestions to improve oncology care by decreasing toxicities while maintaining efficacy and allowing more patients to benefit from treatments for longer periods [5]. Rather than using the MTD as the default approach, ideally, the pre-registrational dose-finding study would be randomized, compare at least two doses, and confirm the dose selected for the registrational trial to see the maximal benefit and the best safety. Continued education will support a realization of the value of these studies in the premarket setting. Sponsors should conduct preclinical research to get sufficient pharmacology data and engage with the FDA to refine and build the dose-finding trial. The overarching goal is that dose-finding studies will be part of standard oncology drug development in the premarket setting to allow delivery of efficacious and tolerable doses to patients at initial marketing approval of a new drug.

Friends suggest the path as studies as below.

The Oncology Center of Excellence (OCE) in FDA has started its initiative called “Project Optimus” to reform the dose optimization and dose selection paradigm in oncology drug development [6, 7]. FDA wants oncology drug developers to determine the adequate dose and schedule for molecularly targeted agents, not based on the dose finding method for cytotoxic chemotherapeutics, before initiating registration trials. When two different doses provide more or less similar efficacy, FDA emphasizes the clinical dose to be chosen through dose optimization study ‘before initiation of phase III trials.’ This concept supports patients’ compliance, maintaining efficacy longer with ultimate benefit of survival.

FDA describes the goals of Project Optimus as follows:

FDA has run a workshop in May 2022 to discuss how non-clinical data can support clinical dose finding, goals of the first-in-human trials, designing to evaluate multi-dose before marketing, safety information beyond the first cycle and PROs [8]. It is expected that FDA would issue guidance for drug developers based on this workshop result.

The accelerated approval of sotorasib (LumakrasTM) announced on May 28, 2021, was based on a study (Codebreak100) of 126 patients with locally advanced or metastatic KRAS G12C-mutated non–small-cell lung cancer (NSCLC) [2]. It was developed with dose escalation with the goal of having a 20–33% incidence of dose-limiting toxicity. The patients in the trial received daily doses of 180, 360, 720, and 960 mg, with two-thirds of the 129 participants receiving the highest dose. The highest dose of 960 mg was judged well tolerated although 58 patients (45%) were classified as having serious complications and nine patients (7%) discontinued treatment because of toxicity. Among patients with platinum-based failure (n = 100), the overall response rate was 36% and the duration of response was 10 months. The submitted evidence meets the statutory evidentiary standard for accelerated approval of sotorasib.

FDA benefit-risk document cited “While the benefit-risk profile in the indicated population is positive based on a response rate that exceeds that of available second-line therapies, the current dose (960 mg QD) may not be optimized, requiring the conduct of a post-marketing dose optimization study.” However, the durable response rate and favorable safety profile observed with sotorasib treatment in a patient population with a significant unmet need and no other available targeted therapies supports approval at this time. Therefore, based on a favorable risk:benefit assessment, sotorasib was granted accelerated approval for the following indication: “Sotorasib is indicated for the treatment of adult patients with KRAS G12C-mutated locally advanced or metastatic non-small cell lung cancer (NSCLC), as determined by an FDA-approved test, who have received at least one prior systemic therapy.”

Upon approval, the FDA has requested that Amgen compare their approved 960-mg dose with a 240-mg dose, as one of their post-marketing requirements. This is the first time the FDA required the labeled dose for targeted oncology drugs to be compared with the lower dose used in phase I trial for confirmation of optimal dose with the similar efficacy and better safety. The FDA review explains they do not consider 960 mg to be the optimal dose for the reasons following: 1) There was no relationship between dose and drug exposure at steady-state, 2) there was no evidence of dose-response relationship, 3) gastrointestinal toxicities may be reduced at a lower dose, 4) preclinical data suggest that the minimal effective dose is 30–240 mg daily, and 5) the labeled dose requires patients to take eight tablets at a time, which is a pill-burden on patients [9].

Drug developers in oncology field, especially for targeted agents and immune-oncology drugs, MTD may not be the right dose, with consideration of a better dose for similar efficacy and better safety. For some drugs with clear evidence of dose-response and exposure-response, the conventional method using MTD may be the right way to pursue. However, in the absence of dose optimization studies mandated by the FDA and other regulatory agencies, the burden of excessive doses would fall on patients, leading to unnecessary toxicities and reduced adherence, and finally less efficacy. With FDA’s stance to demand dose optimization study before pivotal confirmatory trial, it would be necessary to find biologically and clinically optimal dose in early clinical development in oncology. This paradigm change will require extensive education and continuous collaboration of sponsors, investigators, biostatisticians, prescribers, patients, and institutional review boards.