loop through data twice & remove row if duplicate values are found

Very simply, I have a spreadsheet with a column of concatenations in it, listed in chronological order (row 1 being earliest, row x being the last).

In the concatenation column, I would like to seek out the first instance of each unique value ending in A, and delete all subsequent rows with the same value. Then I'd like to cycle back through and do the same thing with rows ending in B. I would prefer if the output were on the same sheet as the input (although I couldn't do that on the sample sheet for obvious reasons).

This is a new approach to a post I made earlier, just boiled down to the theoretical solution instead of the practical solution. I think if this part can be solved, the rest will be easy to record on my own using the macro recorder.

Link to earlier post:
http://www.excelforum.com/excel-prog...ng-on-row.html