Wow! That's ugly!

I see classic fatigue features in the upper right portion, adjacent to the hex region, however, it looks like the fatigue is secondary to a pre existing defect.

Tortional loads will normally cause a spiral type fracture surface, and you sorta have that in the fatigue region (clean area). The long, flat (and dirty) area is very suspect. As was said, that's the older (and more significant) portion of the fracture. Maybe a lamination (or some other planar defect) in the steel from manufacture? Original sin?

Have to also say that pre existing defects (like large laminations) in wrought products is really rare these days, unless you're talking about material sourced from China or India - then all bets are off. But even they are raising the bar compared to some years ago.

If you clean the fracture in a good solvent with a non-metal bristle brush, you may be able to use the concentric beach marks and river patterns to locate an exact origin. Sorry for the fractography garbledegook. It's fairly intuitive.

Too bad you're not closer. If you clean it and take more photographs in macro-mode, I'd be happy to take a look. Or you could just buy new bars and move on with your life....