Why not just have read.csv read in the first header row (which are the actual headers as i understand your question) then skip the next row:
read.csv(file, header=T, skip=1)
Alternatively, if that second header line begins with an idiosyncratic character (not found in your data) then you can specify that line as a comment line by passing in the character that begins the line as the value to the argument comment.char (if that line began w/ "#" for instance, it would be)::
read.csv(file, header=T, comment.char="#")
