There's always more than one way to skin a cat.
Although not terribly manipulation friendly and tedious, you can create a Bitmap or BitmapData object of the frame. Upon clicking you can test the click position as well as every pixel of the object. If the pixel clicked on is transparent (directly with BitmapData or using Bitmap.bitmapData) then you'll know the user intends on clicking through the frame. You then run the code necessary to control the picture.
Outside that your frames may need to be cut up into separate objects (top/left/bottom/right) to separate their hitboxes so the center is genuinely not a clickable transparent area. Also not easy but would allow the click to go fairly directly to the image below.
There's other methods like using the frame as a mask with some pixel detection, yada yada. None of these are nearly as easy as a common sense way to select what you want to edit. If it were flex their components even have a mouseEnabledWhereTransparent property but I haven't seen that outside Flex.