Filtering BOLD signal

Hi all. I have a question related to filtering fMRI signal.

I added all the cosines to my first-level analysis in AFNI (I do not use the CompCor). I performed the first-level analysis with 3ddeconvolve and I included to option polort-A.
My question is : I am double filtering the data? And if not, how do adding the cosines and running polort-A interfere with each other?
One last question: is there a way to also apply a high pass filter equal to 180sec by adding the cosines from fmriprep?

Thanks a lot for your help,