Study design considerations to assess the impact of potential drug–drug interactions in first‐in‐human studies in oncology drug development

First‐in‐human studies are limited to patients with serious diseases for which no curative therapies are available to ensure that the benefits outweigh the risks. However, many patients receive medications that are either victims or perpetrators of drug–drug interactions as part of standard of care. This commentary discusses the current challenges and approaches to safely develop these drugs in patients that require concomitant medications that are potentially either victims or perpetrators of drug–drug interactions.

First‐in‐human (FIH) studies in oncology are usually conducted in patients to provide early access to potentially effective drugs in the absence of available therapies and to identify the recommended dosages for the intended patient population. However, enrollment often restrict intrinsic and extrinsic factors that could impact safety and substantially influence dose finding.

Patients with cancer are treated with concomitant therapies for medical conditions unrelated to their cancer as part of the standard of care for their disease, or to mitigate adverse reactions. ¹ For example, patients with glioblastoma multiform often suffer seizures due to brain metastasis and are treated with anti‐epileptic agents, ² some of which are cytochrome P450 (CYP) inducers; patients with acute myeloid leukemia (AML), receive azole antifungal drugs to prevent infections, ³ many of which are inhibitors of CYP3A. Although these concomitant therapies have been shown to reduce morbidity and mortality, ² , ³ , ⁴ they may interact with oncologic drugs and negatively impact either its safety or activity.

In vitro and in vivo studies ⁵ , ⁶ are essential to identify the potential for interactions between the concomitant therapies and the new oncologic drug before starting clinical studies. The conduct of FIH studies will be challenging if the concomitant therapies that can interact with the new oncologic drug cannot be avoided and limited information is available to understand the effect of the concomitant therapies on drug exposure, safety, and activity of the new oncologic drug. Furthermore, these interactions can confound the selection of the optimal dose for these patients.

This commentary describes the challenges and approaches to developing and identifying safe and active doses for new oncologic drugs when administered with necessary interacting concomitant therapies.

FIH STUDIES IN ALTERNATIVE PATIENT POPULATIONS DUE TO INTERACTING CONCOMITANT THERAPIES

When a new oncologic drug is not expected to be cytotoxic, genotoxic, or target a specific genetic alteration found in only the intended study population, it may be reasonable to conduct FIH studies in healthy subjects or in patients who do not routinely require the interacting concomitant therapies. These data from the FIH study can then be leveraged to select doses to be conducted in a drug‐drug interaction (DDI) study. As such, DDI studies can be conducted in a separate study to help select the recommended dosing regimen of the new oncologic drug for intended patient population when administered with interacting concomitant therapies based on exposure matching.

This approach was used for glasdegib, a Hedgehog pathway inhibitor, approved for the treatment of patients with newly diagnosed AML who are on nonintensive chemotherapy. ⁷ In vitro studies indicated that glasdegib is metabolized by CYP3A4. Initially, dose escalation studies were conducted in patients with hematologic malignancies and solid tumors not requiring azole antifungals. These studies demonstrated a maximum tolerated dosage of 400 mg once daily (q.d.) in patients with hematological malignancies, dose proportional pharmacokinetics (PKs), and saturable pharmacodynamic (PD) activity at 100 mg q.d. ⁷ A DDI study conducted to understand the effects of a strong CYP3A inhibitor on the PKs of glasdegib following a single dose of 200 mg showed that a strong CYP3A4 inhibitor (i.e., ketoconazole) increased glasdegib exposure by twofold. ⁷ Given the saturable PDs observed at 100 mg, and a twofold increase in exposure with a strong CYP3A inhibitor, 100 mg q.d. was selected for the registration trial in patients with AML to provide an adequate safety margin for the increased glasdegib exposure with concomitant use of azole antifungals. ⁷

FIH STUDIES IN THE INTENDED PATIENT POPULATION WHEN THE INTERACTING CONCOMITANT THERAPIES CANNOT BE AVOIDED

FIH studies that need to be conducted in the intended patient population that require interacting concomitant therapies should be conducted with a large safety margin. Although most patients enrolled into these FIH studies will be administered an interacting concomitant therapy, a cohort of patients who will not receive the interacting concomitant therapy should be enrolled for comparative purposes, as they will provide critical PK, efficacy, and safety for the indicated patient population.

This approach is illustrated by the FIH study in patients with relapsed or refractory acute leukemias for SNDX‐5613, a CYP3A4 substrate. ⁸ The study included one arm with patients who required azole antifungals that are CYP3A4 inhibitors and another arm with patients that did not. Dosing for each arm were started at the same dose of 113 mg Q12h. Data from the first two dose levels indicated an increase in SDNX‐5613 exposure with strong CYP3A4 inhibitors of approximately twofold, and higher rates of adverse reactions, including QT prolongation. ⁸

This approach poses several challenges, particularly the possibility of underestimating the risk for potential DDI when selecting a starting dose. It is important that the starting dose administered with the interacting medication have an adequate safety margin to support its use in FIH studies. In addition, different concomitant therapies within the same class (e.g., posaconazole vs. voriconazole) or the same CYP inhibitor in different dosage forms or dosing regimen may have different magnitudes of effect on the exposure of the new oncologic drug. For example, the exposure of ibrutinib varied with different formulations and dosing regimen of posaconazole, and with voriconazole and other strong CYP3A inhibitors. ⁹ As such, early DDI studies or dose escalation studies with and without interfering drugs may not provide a complete assessment of the potential interactions, but can inform on the impact of concomitant medications on the recommended phase II dose during the course of development.

FIH studies in patients who will be treated with a concomitant therapy should consider the following:

• Assess the potential for DDIs based on in vitro metabolic activity screens and/or in vivo DDI studies. These data will inform the selection of a safe starting/dose escalation plan, and/or the choice of medications to address the non‐disease morbidity.

• If proceeding without a separate DDI study, a cohort of patients without interacting concomitant therapy provides a comparison to assess the potential impact of the medications on the investigational drug.

• An appropriate dose escalation strategy to safely identify the recommended dosage(s) for further development of the new oncologic drug when co‐administered with the interacting concomitant therapy should be implemented.

A staggered dose escalation approach may be incorporated into these FIH studies, wherein the safety, PD, and PK data from the initial cohorts are utilized to inform dosing in the interacting arm and subsequent dosing cohorts in both arms.

POTENTIAL FOR EFFECT OF THE NEW ONCOLOGIC DRUG ON CONCOMITANT DRUGS

There are instances when a new oncologic drug is also a perpetrator of the DDI and potentially affect the exposure of the commonly used concomitant therapies, thus impact the safety or effectiveness of the concomitant drugs. Physiologically‐based PK (PBPK) modeling and simulations can be leveraged to understand the impact of potential DDIs and to provide dosage recommendations to be tested during clinical development. An example which highlights this approach is ivosidenib, a kinase inhibitor approved for the treatment of adult patients with AML with a susceptible IDH1 mutation. ¹⁰ Ivosidenib exhibits nonlinear kinetics, is metabolized by CYP3A4, and induces CYP3A4. ¹⁰ Ivosidenib is not only prone to interactions from strong CYP3A4 inhibitors and inducers, but it may in fact also decrease exposure to antifungal drugs that are CYP3A substrates and their effectiveness. The results from an in vivo DDI study showed that concurrent use of a strong CYP3A4 inhibitor (e.g., itraconazole) increased single‐dose ivosidenib exposure by 169%. ¹⁰ This study was used as the basis for PBPK modeling and simulations to predict the effect of ivosidenib as a CYP3A inducer: multiple doses of ivosidenib could reduce the single‐dose exposure of a CYP3A4 sensitive substrate, such as midazolam by 83%, and the steady‐state exposure of itraconazole (also a CYP3A4 substrate) by 90%. ¹⁰ The simulations also predicted an increase in the steady‐state exposure of ivosidenib by 3.8‐fold with a strong CYP3A4 inhibitor which itself is not a substrate of CYP3A4. Further, the PBPK simulations indicated a 1.9‐fold increase in ivosidenib with a moderate CYP3A4 inhibitor (e.g., fluconazole). ¹⁰ The PBPK modeling formed the basis for the ivosidenib labeling recommendations to healthcare providers to avoid co‐administration of azole antifungals that are CYP3A4 substrates, including itraconazole and ketoconazole, as it may result in a loss of antifungal efficacy, and to reduce the dose of ivosidenib when co‐administered with a strong CYP3A inhibitor as it may increase exposure of ivosidenib and potentially increasing the severity and incidence of adverse reactions, including QT interval prolongation. ¹⁰

CONCLUSION

During drug development, it is critically important to identify the optimal doses of the new therapeutic for patients to assure safety and efficacy. For drugs that may be impacted by their concomitant medications, nonclinical and in vitro studies are key to understanding the potential risks for DDIs with a new oncologic drug and designing FIH studies to select doses that account for these DDI issues. Strategies, such as exploring the DDIs in healthy subjects or in patients with other cancers who do not require interacting concomitant therapies, where possible, or evaluating the drug in the intended patient population with additional strategies to safely evaluate the PKs, safety, and activity of the new oncology drug, have been successfully used to identify recommended dosage(s) for further clinical development and support labeling recommendations. These approaches help minimize exposure to toxic or ineffective dosages and help identify the recommended dosage(s) for the intended population who will be receiving interacting concomitant therapies.

FUNDING INFORMATION

No funding was received for this work.

CONFLICT OF INTEREST STATEMENT

The authors declared no competing interests for this work.

DISCLAIMER

The opinions expressed in this manuscript are those of the authors and should not be interpreted as the position of the US Food and Drug Administration.

Subramaniam S, Shord SS, Leong R, et al. Study design considerations to assess the impact of potential drug–drug interactions in first‐in‐human studies in oncology drug development. Clin Transl Sci. 2023;16:719‐722. doi: 10.1111/cts.13496