[Clinical Tumor Research] Issues to be considered in PFS analysis

In clinical oncology research, disease-free survival (DFS), objective response rate (ORR), time to disease progression (TTP), progression-free survival (PFS) and time to treatment failure (TTF) are all trials based on tumor measurement End points, and the data collection and processing of the above-mentioned time-dependent end points are based on indirect evaluation, calculation or estimation (such as tumor measurement)
.

Among the above indicators, PFS is a commonly used endpoint in cancer clinical trials, and its definition is the time from the start of randomization to the appearance of objective tumor progression or death
.

The sister platform of "NEJM Frontiers in Medicine" praised Tumor Monday's “Clinical Tumor Research Column” and shared issues related to PFS analysis with you
.

PFS as an end point for supporting drug approval Table 1 lists the advantages and disadvantages of PFS as an end point for supporting anti-cancer drug approval
.

PFS reflects the growth of the tumor, and can be evaluated before the survival benefit is confirmed, and will not be affected by subsequent treatment
.

However, for a certain sample size, the impact of various factors on PFS may be greater than the impact on overall survival
.

Table 1.
Advantages and disadvantages of PFS as an end point to support the approval of anticancer drugs.
However, for many different types of malignant tumors, it is difficult to formally confirm PFS as a surrogate end point for survival
.

There are usually insufficient data to evaluate the correlation between survival and PFS
.

The scale of clinical trials of anti-tumor drugs is usually small, and the survival benefits of existing drugs are usually small
.

In different anti-tumor treatment trials, the PFS endpoints used to support approval play different roles
.

Whether the extension of PFS directly represents clinical benefit or just a surrogate end point for clinical benefit depends on the efficacy of this new treatment method and the risk-benefit ratio compared with existing treatments
.

The experimental design issues should describe in detail the methodology for evaluating, measuring, and analyzing PFS in the experimental protocol and statistical analysis plan
.

Similarly, it is also important to define the tumor progression criteria in detail in the trial protocol
.

At present, there is no standard for the definition of tumor progression, and researchers may use different standards, including the RECIST standard
.

In the two experimental groups, follow-up and imaging evaluation must be balanced to avoid systematic bias
.

Research should be blinded as much as possible.
If patient or investigator evaluation is a factor in the progress endpoint, blinding is particularly important.
Physician) for evaluation
.

If the research is for new drug marketing approval, the regulatory agency should reach an agreement with the investigator in advance on the following aspects: research design; definition of disease progression; data recorded on the clinical trial observation form (CRF); statistical analysis plan; Methods of handling missing data and data censorship; if applicable, the operating procedures of the Independent Endpoint Review Committee (IRC)
.

PFS analysis due to lack of data, PFS analysis will face some difficulties
.

In this case, the work that needs to be done by the investigator includes: The definition of "sufficient evaluation follow-up (that is, all agreed tumor evaluations completed during this follow-up)" should be given to each patient in the trial protocol; The analysis plan should summarize and compare the adequacy of follow-up of each treatment group; the plan should specify how to analyze incomplete and/or missing follow-up data and the method of data censorship; the analysis plan should clearly specify the main analysis and one or more sensitive Analysis to evaluate the reliability of the results
.

When collecting data over a longer period of time, it is often difficult to determine the date of the event and the date of censorship
.

In this case, if there is no missing evaluation before, the earliest observation of disease progression is marked as the progress date, and the date of the last imaging evaluation at which no progress is determined is determined as the cut-off date
.

In addition, through one or two secondary analysis to evaluate the expected problems in the clinical trial and whether the evaluation results are reliable
.

Here are some important factors to consider
.

Definition of the date of progression Survival analysis uses the exact date of death for analysis
.

However, in PFS analysis, the exact date of progress is often unknown
.

The following are two methods used in PFS analysis to define the recorded progress date (PDate)
.

1.
Set PDate as the first time to prove progress
.

Taking the appearance of new lesions as a criterion for progress, PDate refers to the date when a measurable new lesion was first observed
.

If multiple evaluations based on the sum of the measured values ​​of the target lesion are performed at different times, PDate refers to the date of the last observation or imaging examination of the target lesion, which can show that the increase in the sum of the measured values ​​of the target lesion reaches a predetermined threshold
.

2.
Define PDate as the date of clinical follow-up as planned after all imaging examinations are completed (comprehensive proof of progress)
.

Definition of Censoring/Censoring Date Censoring applies to patients who have no progress records before data cutoff or withdrawal
.

For these patients, the censorship date is defined as the final date when there is sufficient evidence of progress
.

One feasible method is to use the date of the final evaluation, or multiple imaging studies can be used to determine progress
.

Another feasible method is to use the clinical follow-up date corresponding to the imaging examination
.

The definition of adequate tumor evaluation For patients with no evidence of disease progression, the censorship of PFS usually depends on the date of the last adequate evaluation of the tumor
.

The precise definition of adequate tumor evaluation includes adequate evaluation of target lesions and adequate imaging tests for evaluating non-target lesions and finding new lesions
.

Analysis of Partially Missing Tumor Data When there is a partial missing of data during the follow-up period of full evaluation of the tumor, the analysis plan should describe the method for calculating the progress status
.

It is impossible to measure the progress of the disease.
If possible, the progress criteria should be described for each method of evaluating morphological characteristics (eg CT scan, bone scan)
.

Through a blinded review committee, it is very important to check the diseases that cannot be measured but the progress of imaging is confirmed, and to be verified by the National Medical Products Administration
.

During the analysis of suspicious lesions, calculation methods for evaluation and follow-up should be provided for unclear lesions for the specified progress state
.

The endpoints of any clinical trial have both advantages and disadvantages
.

As a commonly used endpoint, the researcher should fully consider and determine a reasonable experimental design and analysis method before the start of the experiment, so as to truly achieve the experimental goal
.

*This article is adapted from the "Technical Guidelines for Endpoints of Clinical Trials of Anti-tumor Drugs" issued by the State Drug Administration.
For the original text, please refer to the column of guidelines for the Center for Drug Evaluation of the State Drug Administration
.

Copyright information This article is written or commissioned by the editorial department of Jiayi Oncology
.

If you need to reprint, please contact nicole.
liu@j-med.
com
.

Unauthorized use of article content and part of the content is an infringement, and the copyright owner reserves the right to pursue legal liabilities
.