How to censor TCGA survial data for a specific number of days?

Question

[akhattri]

We are analyzing some TCGA Overall Survival (OS) and Progression Free Interval (PFI) data. As for some cancer types the OS could be very long as compared to others. Moreover, with a longer duration, the probability of events (death) due to unrelated reasons (like, due to a car accident) increases, and hence several studies censor the data for five years (1825 days).

I am not sure of the correct censoring strategy for OS and PFI data. Do we consider all the numbers >=1825 as 1825 days or is there any other thing to consider? I am providing a small subset of the actual TCGA data that we would like to analyze. Thanks!!

ID      type    OS  OS.time PFI PFI.time
B6-A0IA brca    0   8391    0   8391
B6-A0RN brca    0   8008    0   8008
B6-A0RE brca    0   7777    0   7777
BH.A18T brca    1   0224    1   224
AN.A0FN brca    0   0218    0   218
B6.A400 brca    0   0215    0   215
AN.A0FK brca    0   0213    0   213