echemi logo
Product
  • Product
  • Supplier
  • Inquiry
    Home > Active Ingredient News > Antitumor Therapy > Pre-researcher must read Phase I dose exploratory design considering multiple toxicity levels or multiple toxicity types: MT-BOIN

    Pre-researcher must read Phase I dose exploratory design considering multiple toxicity levels or multiple toxicity types: MT-BOIN

    • Last Update: 2021-06-11
    • Source: Internet
    • Author: User
    Search more information of high quality chemicals, good prices and reliable suppliers, visit www.echemi.com
    Author: Zhang Jiwei This article is published by Yimaitong authorized by the author, please do not reprint without authorization.

    In the phase I dose exploration tumor clinical trial, our main purpose is to find the human body’s maximum tolerated dose (MTD) or optimal biological dose (OBD, suitable for molecular targeted therapy and immunotherapy drugs), Provide a theoretical basis for the recommended dose for phase II administration.

    In order to achieve this goal, our consistent approach is based on the number of subjects enrolled in the group according to the current dose level and the number of subjects with dose-limiting toxicity (DLT) (or the number of subjects with dose-limiting toxicity).
    Incidence rate), combined with the dose given in the plan to explore the design method and its dose increase and decrease rules (such as traditional 3+3 design, BOIN design, mTPI design, Keyboard design, CRM design, etc.
    ), in the set dose level Find the maximum tolerated dose or the best biological dose.

    There are two "number of subjects": one is the number of subjects enrolled in the group at the current dose level, which can be directly observed during the experiment; the other is the number of subjects with dose-limiting toxicity at the current dose level The number of patients, which needs to be judged by the investigator through the concept of dose-limiting toxicity defined in the protocol.

    That is, if a certain toxicity event occurs (for example, hematological toxicity with toxicity level ≥ 3), it is considered that there is a dose-limiting toxicity, and the corresponding number of subjects is increased by 1; otherwise (for example, hematology with toxicity level <3 Toxicity) is considered to have no dose-limiting toxicity.

    When calculating the number of dose-limiting toxicity subjects, there are two conditions for attention: one is to use the toxicity event as the secondary end point ("Yes" or "No") to indicate whether the subject has a dose Restricted toxicity; the second is to classify different types of toxic events into a category of endpoints, such as the classification of toxic events according to the National Cancer Institute Common Toxicity Criteria for Adverse Events (NCI-CTCAE) Subjects with grade 3 thrombocytopenia and grade 3 neutrophil count reduction may be classified as a dose-limiting toxicity.

    Based on the above two application conditions, we will have the following thinking: Thinking 1: Are the above two conditions reasonable and whether there are other circumstances? We all know that in clinical trials, toxic events are classified into 5 grades according to NCI-CTCAE.
    If a binary end point is used to aggregate 5 grades of toxic events, a lot of information will be lost.

    Especially in molecular targeted therapy and immunotherapy, the level of toxic events is often relatively low.
    At this time, it is particularly important to pay attention to low-level toxic events in dose exploration.

    In addition, categorizing different types of toxic events into one category will also lose a lot of information.

    In many cases, paying attention to different types of toxic events is also particularly important for accurately exploring MTD.

    Thinking 2: If considering multiple toxicity levels or different toxicity types, is there a new experimental design specifically to deal with this situation? Many scholars have proposed different design methods in phase I dose exploration clinical trials considering multiple toxicity levels or different types of toxicity.

    For example, Bekele and Thall proposed that the toxicity levels of multiple different types of toxic events were weighted and summed to construct a TTB (Total Toxicity Burden) indicator to explore MTD; Yuan et al.
    used quasi-CRM method to explore MTD; Ezzalfnai et al.
    constructed TTP (Total Toxicity Profile) indicators to explore MTD; the gBOIN design proposed by Mu et al. However, these designs finally classify multiple toxicity levels or multiple toxicity types into a single indicator through various statistical methods, thus losing the original level of toxic events, which also loses a lot of information.

    In order to fully consider the original level of toxic events when exploring MTD, Lee et al.
    proposed the MC-CRM design.

    However, the design is based on the CRM design framework, so it needs to build a complex model and requires a relatively high level of computer.

    Today, I will introduce a new experimental design method, MT-BOIN design (Multiple-toxicity Bayesian Optimal INterval design).
    Compared with the method mentioned above, this design is convenient to operate and easy to implement, and the decision rule of dose increase and decrease is the same as BOIN.
    The design can be specified in advance before the test starts.

    The most important thing is that the design can consider multiple toxicity levels or types of toxicity.

    What is MT-BOIN design? This design is a method published by Professor Ruitao Liu of MD Anderson Cancer Center in the United States in The International Biometric Society in 2018.
    It is mainly used to solve the phase I dose exploration clinical trial when multiple toxicity levels or types of toxicity are considered.
    .

    The principle of MT-BOIN design: transform the dose exploration problem under multiple toxicity levels or types of toxicity into a Bayesian decision problem, by minimizing the maximum joint probability of making wrong decisions, finding the boundary value of dose increase and decrease, and then according to The corresponding increase and decrease rules for dose exploration.

    Finally, order-preserving regression is used to determine MTD.

    Types of MT-BOIN design: Professor Ruitao Liu mentioned three types of toxicity of MT-BOIN design in the article, which are as follows: The first type: toxicity results of equal importance.

    For example, when exploring MTD, it is necessary to consider the toxicity of a drug on multiple organs at the same time.
    Assume that there are two toxicity results Y1 (representing the toxicity of the drug to the liver) and Y2 (representing the toxicity of the drug to other organs).
    At this time, we Look at the two toxicity results Y1 and Y2 equally. The second type: toxicity results of unequal importance.

    In some dose exploration trials, the importance of the toxicity results that people are concerned about may vary.

    For example, the toxicity result Y1 (representing death related to the study drug) is our most concern, while the toxicity result Y2 (other toxicity events related to the study drug except death) is of secondary concern.

    At this time, Y1 and Y2 are toxic results with unequal importance.

    In this case, the decision to increase or decrease the dose is mainly based on the main toxicity result, and only if the main toxicity result is well controlled, should the secondary toxicity result be considered.

    The third type: nested toxicity results.

    For example, a test defines two toxicity results Y1 and Y2 based on NCI-CTCAE.
    Y1 is defined as a toxicity event with a toxicity level ≥ 4, and Y2 is defined as a toxicity event with a toxicity level ≥ 3.

    At this time, Y1 and Y2 are called nested toxicity types.

    MT-BOIN design operation steps: the first step: according to the multiple toxicity grades or toxicity types that need to be considered in the plan, and the set target toxicity probability value, calculate the decision rule table for dose increase or decrease, as shown in Figure 1.
    Shown: Figure 1 Toxicity end point is Y1 (target toxicity probability is φ1=0.
    1) and Y2 (target toxicity probability is φ2=0.
    25) MT-BOIN design dose increase or decrease decision rule table Step 2: According to different toxicity types, select the corresponding The dose increase and decrease rules.

    2a) The rules for increasing and decreasing the dose of toxic results of equal importance are as follows: ① Start enrolling subjects from the lowest dose level or the dose level specified by the researcher.

    ② According to the number of subjects nj enrolled in the group according to the current dose level j and the number of subjects with dose-limiting toxicity yj, combined with the dose increase or decrease decision rule in Figure 1, proceed as follows.

    ③ Repeat the above steps until the maximum sample size specified in the protocol is reached, or the safety margin set in advance is reached (for example, each dose level group enrolls 12 more subjects, or based on the current dose level data, It is estimated that the probability of its toxicity probability is greater than the target toxicity probability by more than 95%, etc.
    ), the test can be stopped. ④ After the test is completed, the order-preserving regression is used to estimate the maximum tolerated dose MTD.

    2b) The rules for increasing or decreasing the dose of toxic results of unequal importance are as follows: ① Start enrolling subjects from the lowest dose level or the dose level specified by the investigator.

    ② According to the number of subjects nj enrolled in the group according to the current dose level j and the number of subjects with dose-limiting toxicity yj, combined with the dose increase or decrease decision rule in Figure 1, proceed as follows.

    ③ Repeat the above steps until the maximum sample size specified in the plan is reached, or the safety limit set in advance is reached, the test can be stopped.

    ④ After the test is completed, the order-preserving regression is used to estimate the maximum tolerated dose MTD.

    2c) The dose increase and decrease rules for nested toxicity results are similar to the above.
    Interested readers can check the literature on their own.
    Due to the limited space, we will not repeat them here.

    The above briefly shared the basic theory and related operating steps of the MT-BOIN design method proposed for the phase I dose exploration trial design when considering multiple toxicity levels or types of toxicity.

    Due to my limited level and understanding, if there is something wrong, you are welcome to criticize and correct me.

    Setting a fixed width and height background on the fixed layout toolbar can be set to be included, which can perfectly align the background image and text and make your own templates.
    Reference Lin R.
    Bayesian optimal interval design with multiple toxicity constraints.
    Biometrics.
    2018 Dec;74(4): 1320-1330.
    doi: 10.
    1111/biom.
    12912.
    Epub 2018 Jun 5.
    PMID: 29870069.
    Recommended reading 1.
    Let every clinical researcher do drug combination trial design | software demonstration 2.
    Clinical trial things | drug combination therapy Trial BOIN Comb design and application skills: find a single MTD3.
    Clinical trials of those things | I think about drug combination therapy in the phase I dose exploration trial 4.
    Clinical trials of those things | Application of Waterfall design in drug combination trials: find Multiple MTD5.
    Case sharing: How to use Waterfall design in co-administration trials 6.
    Pre-researchers must read | Phase I dose exploration design considering binary or continuous toxicity endpoints: gBOIN7.
    Pre-researchers must read | Consider history Phase I dose exploration design based on data and real-world evidence: iBOIN8.
    A must-read for temporary researchers | Software demo: teach you how to operate iBOIN design
    This article is an English version of an article which is originally in the Chinese language on echemi.com and is provided for information purposes only. This website makes no representation or warranty of any kind, either expressed or implied, as to the accuracy, completeness ownership or reliability of the article or any translations thereof. If you have any concerns or complaints relating to the article, please send an email, providing a detailed description of the concern or complaint, to service@echemi.com. A staff member will contact you within 5 working days. Once verified, infringing content will be removed immediately.

    Contact Us

    The source of this page with content of products and services is from Internet, which doesn't represent ECHEMI's opinion. If you have any queries, please write to service@echemi.com. It will be replied within 5 days.

    Moreover, if you find any instances of plagiarism from the page, please send email to service@echemi.com with relevant evidence.