What is the difference between OS, PFS and DFS?

When it comes to cancer clinical research, we have to talk about the end points of clinical trials, such as the well-known OS, PFS, ORR, DFS, TTP, TTF.

Now let us take a look at the differences, advantages and disadvantages of commonly used clinical trial endpoints

Overall survival

overall survival, OS

Defined as: the time from the start of randomization to death (for any reason)

OS is considered to be the best efficacy endpoint in cancer clinical trials, and it is the preferred endpoint when the patient’s survival can be fully assessed

The biggest advantage is that it is convenient to record.

There are also shortcomings.

After treatment, some tumor patients recurred and metastasized, some died, and some survived

Of course, three-year survival rate and ten-year survival rate are also useful to express curative effects

Except for OS, the rest of the endpoints are based on tumor measurements

Different tumor trials have large differences in the accuracy of tumor measurement, which requires researchers to fully evaluate the benefits and biases

Objective response rate

objective response rate, ORR

Defined as: the proportion of patients whose tumor volume has shrunk to a predetermined value and can maintain the minimum time limit, which is the sum of the proportions of complete remission and partial remission

The remission period usually refers to the period from the onset of efficacy to the confirmation of tumor progression

The ORR remission criteria should be defined in advance in the protocol before the start of the trial.

Progression-free survival

progression-free survival, PFS

Defined as: the time from the start of randomization to the progression of tumors (in any aspect) or death (for any reason)

Compared with OS, the node "deterioration occurs" is increased, and "degeneration" is often earlier than death, so PFS is often shorter than OS, but it can also be evaluated before OS, so the follow-up time is shorter

And because of the addition of the node "deterioration", the definition of different tumor progression is different, and different studies are prone to bias when judging tumor progression

Therefore, in the design of clinical trials, the criteria for "tumor progression" must be clearly defined, including PFS assessment, observation, and analysis methods.
Follow-up and imaging evaluation must be balanced.
It is best to have an imaging expert A blind independent adjudication team composed of clinical experts was conducted
.

PFS includes death, which better reflects the side effects of drugs, and therefore has a better correlation with OS
.
However, if in the process of evaluating PFS, it is found that most patients died of other diseases instead of tumors, then PFS will inevitably be biased
.
At this time, I have to say another evaluation indicator similar to PFS-TTP
.

Disease progression time

time to progress, TTP

Defined as: the time from the start of randomization to the progression of tumorigenesis (in any aspect)
.

TTP mainly records disease "deterioration" and does not include "death".
It considers tumor activity.
Therefore, when most deaths are not related to tumors, TTP is an acceptable endpoint
.
In addition, if there are multiple treatments, there is a crossover effect-the difference in TTP will not be masked by the second treatment
.

TTP, like PFS, requires a small sample size for evaluation, and the follow-up time is shorter than that of OS
.

The problem is that if the subject has died before worsening, his TTP must not be observed.
At this time, the recorded TTP is incomplete, and it is statistically called censoring.
This is for missing data.
It is difficult to determine the deadline for processing and data
.

In addition, since most clinical trials are not double-blind, this introduces bias into TTP decision-making
.
However, the follow-up of patients is difficult: it is necessary to determine the lesions of all parts, the follow-up time and interval are different, there will be differences in TTP, and it is difficult to determine the clinical significance of the difference
.

It can be seen that TTP is inferior to PFS in predicting clinical benefits, and there are many problems, and a clear definition and evaluation of "progress" is also needed
.

Disease-free survival

disease-free survival, DFS

Defined as: the time from the start of randomization to the recurrence of the disease or death (for any reason)
.

DFS is most commonly used in the research of adjuvant therapy after radical surgery or radiotherapy.
It is currently used as the main approval basis for adjuvant hormone therapy for breast cancer, adjuvant therapy for colon cancer, and adjuvant chemotherapy for breast cancer
.

Recurrence of the disease requires careful follow-up, and recording is also difficult, and cancer patients often have comorbidities that can easily interfere with the judgment of DFS
.
When the patient died outside the hospital, there was no pre-recorded tumor progression, and autopsy was often not possible at this time, and the recurrence could not be determined
.

Treatment failure time

time to treatment failure, TTF

It is defined as: a randomized begin to "exit test" to exit the reason may be that the patient refused, disease progression and death in patients with adverse events
.

Since it does not only show the efficacy of drugs, it is not recommended for efficacy confirmation tests
.

The essence of TTF is an index with comprehensive characteristics, so it can potentially affect the expected curative effect in order to reduce toxicity
.