PD-1/PD-L1

I. Overview

Programmed Cell Death Receptor-1 (PD-1) is an important immune checkpoint that regulates T-cell activity, thereby modulating the body's immune response to its own cells and foreign substances. Programmed Cell Death-Ligand 1 (PD-L1) binds to PD-1 on the surface of T cells, initiating programmed cell death (apoptosis) of T cells, thus inhibiting the T-cell immune response and leading to immune evasion by tumor cells. Anti-PD-1/PD-L1 antibodies can block the PD-1/PD-L1 signaling pathway and have demonstrated good anti-tumor efficacy in cancer patients. They have been developed for the treatment of various cancers.

Without changing the approved indications and routes of administration, for anti-PD-1/PD-L1 antibodies that exhibit a relatively flat exposure-response (E-R) relationship within the therapeutic dose range, alternative dosing regimens can be considered during the development process or after market approval. This may involve changes in dosing (e.g., weight/body surface area-based dosing, stratified dosing based on different factors, etc.) and/or dosing intervals. The development of alternative dosing regimens can facilitate individualized patient treatment, improve medication adherence, reduce risks associated with frequent dosing (such as infusion reactions), and lessen the inconvenience caused by frequent hospital visits. Since alternative regimens differ from the dosing regimen used in the clinical safety and efficacy trials supporting initial market approval, or from regimens used in early pre-market studies, a scientific assessment of the rationale for the alternative regimen is required. The development of alternative dosing regimens can be conducted using clinical trial methods or based on Pharmacokinetic (PK) modeling and simulation approaches.

This guidance primarily elaborates on the modeling and simulation-based method to support the interchange between pre-market and/or post-market dosing regimens for anti-PD-1/PD-L1 antibodies. This method is applicable to PD-1/PD-L1 monotherapy regimens, or combination therapy regimens where only the dose and/or dosing interval of the PD-1/PD-L1 component is changed.

The content of this guidance is primarily based on the current understanding of anti-PD-1/PD-L1 antibodies and the accumulation of safety and efficacy experience, providing relevant considerations and general scientific guidance. As the field continues to develop, further research and analysis should be conducted based on scientific judgment in the future.

II. Research Methodology

(I) Population Pharmacokinetic Modeling

Population PK analysis methods are generally employed for modeling. Based on the established model, the PK characteristics of different dosing regimens are simulated to support alternative dosing regimens for anti-PD-1/PD-L1 antibodies. When simulating and predicting concentration-time curves and PK parameters for different regimens, PK parameters relevant to clinical safety and efficacy should be selected, such as AUC (or C-average), Cmax, and Ctrough (or Cmin). Specific requirements for the population PK model can be found in the "Technical Guideline for Population Pharmacokinetics Research."

(II) Dose/Exposure-Response Relationship Analysis

Dose/Exposure-Response (D/E-R) relationship analysis helps predict clinical response under different dosing regimens or exposure levels, which can support alternative dosing regimens for anti-PD-1/PD-L1 antibodies. Data used for D/E-R analysis should be sufficient and reliable. Specific requirements for D/E-R relationship analysis can be found in relevant guidelines.

III. General Requirements

(I) Acceptance Criteria

An alternative dosing regimen supported by population PK modeling and simulation should possess the following characteristics:

1. For marketed drugs, the reference regimen should be the dosing regimen (including dose and dosing interval/frequency) used in the clinical safety and efficacy trials supporting drug approval. For unmarketed drugs, the reference regimen refers to the regimen used in previous clinical development to characterize the drug's PK, efficacy, and safety.

2. Compared to the reference regimen, the geometric mean reduction in AUC (or C-average) and Ctrough (or Cmin) for the alternative regimen, at steady state and/or over the first minimum common dosing interval (e.g., the minimum common interval for a q2w and a q3w regimen is 6 weeks), should not exceed 20%. Compared to the reference regimen, the geometric mean increase in steady-state Cmax for the alternative regimen should not exceed 25%, unless sufficient clinical evidence demonstrates that the safety risk at the steady-state Cmax of the alternative regimen is acceptable. This could include evidence of safety at higher doses, or evidence of a flat D/E-R relationship within the therapeutic exposure range of the drug.

If the above conditions are not met, additional clinical trial data are required to support the efficacy and safety of the alternative dosing regimen. The necessary clinical trials may depend on the drug's characteristics, patient population, obtained clinical efficacy and safety data, and clinical pharmacology data. Specific issues can be discussed with the review department.

(II) Considerations in Model Development and Simulation

1. Based on Population Pharmacokinetic Models

Since the PK parameter simulation results for the dosing regimens are obtained from the population PK model, the PK data used for modeling should be adequate. The analysis data for the population PK model should encompass all available PK data, including data from various dosing regimens and different patient populations across indications, and should be updated promptly as research progresses. The population PK model should be adequately validated, demonstrating acceptable robustness and reliability. If factors such as tumor type and/or concomitant therapies significantly impact PK parameters, simulations should be performed separately for the target population.

When converting from weight-based dosing to fixed dosing (stratified or unstratified), patients with low/high body weight may experience a shift towards high/low exposure. For instance, drug exposure in low-weight patients may become too high, while exposure in high-weight patients may become too low. Therefore, it is recommended to simulate exposure for the alternative regimen across different body weights (especially extreme weight values included in clinical trials). This simulation should compare exposure differences between the alternative and reference regimens under different weight stratifications, ensuring the safety and efficacy of the expected exposure range for all target body weight populations receiving the proposed dose are acceptable. Additionally, it is advisable to comprehensively consider factors like drug economics, convenience, and adherence. For extreme body weight populations, weight-based dosing may sometimes be more appropriate.

If changing the dosing interval, for example, from every 2 weeks to every 4 weeks, the drug's PK profile will change, with parameters like Cmax and AUC potentially being altered. When assessing the reasonableness of the exposure range for the alternative regimen, the impact of PK profile changes on efficacy and safety must be considered.

2. Based on Dose/Exposure-Response Relationship Models

When analyzing the D/E-R relationship, clinical responses related to both efficacy and safety should be included. To ensure model stability and reliability, the D/E-R relationship model should incorporate a sufficient number of patients from the target indication population and cover a relatively wide dose/exposure range. For some anti-PD-1/PD-L1 antibodies, treatment effects can influence antibody clearance, often resulting in time-dependent clearance. Using steady-state exposure data may reveal a more pronounced E-R relationship. Therefore, in addition to steady-state PK parameters, it is recommended to also examine PK parameters over the first minimum common dosing interval as indicators for D/E-R relationship analysis. When using other PK parameters as indicators in D/E-R analysis, the reliability of this analysis should be fully justified. Data analysis across multiple dose levels can more accurately reflect the D/E-R relationship. E-R analysis based on a single dose level may be more significantly affected by target-mediated clearance, disease status, and their interaction, and the exposure range is relatively narrow. If the E-R relationship is derived from a single dose level, caution should be exercised in its interpretation and application. Efficacy E-R analysis typically incorporates data from studies in the target indication population. Safety E-R analysis often pools data from studies across different indication populations for integrated analysis.

Data for D/E-R analysis can include plasma drug concentration, minimum effective concentration, receptor occupancy, clinical efficacy endpoints (such as overall survival, objective response rate, progression-free survival, etc.), and safety endpoints (such as adverse event incidence). Factors that may introduce variability into the analysis results should be noted, for example, the type of in vitro assays and parameter settings used to measure pharmacodynamic (PD) biomarkers and their correlation with clinical efficacy endpoints.

(III) Other Supporting Evidence

Clinical trials can directly demonstrate the reasonableness of an alternative dosing regimen. Furthermore, the results of PK comparative trials between the alternative and reference regimens can provide exposure differences, supporting the regimen conversion. If partial safety and/or efficacy data for the alternative regimen exist, combined with modeling and simulation analysis results, they can also be used to support the regimen conversion.

During the early development stages of anti-PD-1/PD-L1 drugs, studies on targets or biomarkers are often conducted. When evaluating alternative dosing regimens, relevant information such as receptor occupancy and biomarker indicators under the alternative regimen can also be provided as supportive evidence.

IV. Communication and Interaction

If planning to use PK modeling and simulation methods to support an alternative dosing regimen, applicants are encouraged to communicate with the regulatory agency as early as possible. Meeting materials for communication should include the following information:

1. Background information on the drug: Including the anticipated changes to the dosing regimen, a summary of all supporting data (e.g., receptor occupancy, biomarkers, minimum effective concentration, etc.), and a summary of the development strategy and research methods.

2. Summary information on safety and efficacy E-R relationships: Including E-R relationships for safety and efficacy; E-R relationship analysis for biomarkers supporting the IND application and approval; and the establishment and validation of the population PK model.

3. Simulation strategy and plan: Including study objectives, assumptions, target population, simulation scenarios, E-R model parameters, and data analysis plans.

4. Specific questions related to the modeling and simulation strategy.

V. Submission Materials

Submission materials for an alternative dosing regimen should include all relevant study reports as required by regulations. If clinical trials for the alternative regimen have been conducted, detailed study reports should also be submitted according to regulatory requirements.

1. A comprehensive summary report should be submitted, explaining the purpose of the submission, changes to the prescribing information, relevant information on target or biomarker E-R relationships (if any), summaries of efficacy and safety E-R relationships and results, etc.

2. Complete PK study reports, population PK study reports, model simulation reports, D/E-R relationship analysis reports, etc., should be submitted. The aforementioned studies (if conducted) should comply with domestic and international relevant regulations and guidelines, providing complete raw data and program code, modeling details, model evaluation, and simulation results.

o PK study reports should summarize PK results for different dosing regimens.

o Population PK study reports should provide complete information on model validation and performance assessment across different indication patient populations and dose levels.

o Simulation reports should provide the simulation strategy and its results related to the alternative dosing regimen. Direct comparison results of PK profiles and parameters between the alternative and reference regimens are required. Comparison of PK parameters can be characterized using graphical and numerical analyses. Graphical analysis may include box plots, etc., and numerical analysis may include similarity comparisons, comparisons of the proportion of simulated patients above and below efficacy or safety thresholds, etc.

VI. References

1. FDA. Pharmacokinetic-Based Criteria for Supporting Alternative Dosing Regimens of Programmed Cell Death Receptor-1 (PD-1) or Programmed Cell Death-Ligand 1 (PD-L1) Blocking Antibodies for Treatment of Patients with Cancer. Dec 2022.

2. National Medical Products Administration. "Technical Guideline for Population Pharmacokinetics Research." December 2020.

3. National Medical Products Administration. "Technical Guideline for Model-Informed Drug Development." December 2020.

4. PMDA. Guideline for Exposure-Response Analysis of Drugs. Jun 2020.

5. FDA. Exposure-Response Relationships — Study Design, Data Analysis, and Regulatory Applications. Apr 2003.

6. ICH. E4 Dose-Response Information to Support Drug Registration. Mar 1994.