Merge duplicate values in columns

Given a file like this

value,value,value,value
value1,value1,value,value1
value2,value2,value,value2

How can I transform it to look like this:

value,value,value,value
value1,value1,    ,value1
value2,value2,    ,value2

Basically, to merge duplicate values in column 3 and present it in first row and retain other data, also make this as one record within csv. I have tried cat file | sort -u -t, -k3 but it does not work.