How to do Disease-Free survival analysis with TCGA data ??

Question

[mario.red8976]

Hi everybody! I am trying to do survival analysis with TCGA-PRAD data. I learned how to do overall survival, KM plots and cox regression following this tutorial: https://bioconnector.github.io/workshops/r-survival.html , in which you learn how to use R packages "survival", "survminer" and TCGA data using the package "RTCGA".

My aim now is to do disease-free survival, but looking at the clinical data that i can download with RTCGA, I can't understand which columns I have to use, I can't see a clear indication of dfs. Also, checking here on biostars it seems that there are no other questions related specifically to it.

Can someone please help me to find a way to do this? Some example code will be really appreciated! Thank you a lot!!

[weiqiao1029]

Hi @miky.zo ! Have you solved your problem?I now met the same question with you.I downloaded BLCA clinical data via TCGA-Assembler but I don't know how to get DFS(disease-free survival). I only know the OS(overall survival) is merge days_to_last_followup and days_to_death but have no idea about anything like progress/recurrent. Besides, after I read @MatthewP 's reply I am confused why the DFS in TCGA and cbioportor are different?

[Ram]

Why did you delete this and other posts of yours, mario.red8976 ?

Answer 1

[MatthewP]

Hello, I don't know RTCGA.
The events of DFS is progress/recurrent so you need to find which column contains this information. I personally suggest you download clinical data from cBioportal, and you will get 2 columns Disease Free (Months) and Disease Free Status.
BTY, I did DFS for TCGA-PRAD before and it seems too little event cases. You can try.

[mario.red8976]

Hi, thank you for your answer! The problem is that in PRAD, there are very few death events (like 10 in 500 cases) so I don't think that doing overall survival will give good results with this dataset, so I thought that DFS might be better.

[MatthewP]

DFS also too few events in TCGA-PRAD. PFS(Progress Free Survival) has many more events in clinical data from cBioportal.

[mario.red8976]

Ok, I'll try pfs, thank you!

[weiqiao1029]

why the DFS is different in TCGA and cbioportal?

[MatthewP]

Because clinical data of TCGA has different release.