This should be what you need:
(?<Num>[\d\s]*)\k<Num>(\d*)\s*(\d*)\s*(\d*)\s*(\d*)
This expands on Michael's answer to remove the spaces, but also removes the decimal and anything following the decimal, which I believe it what you were looking to do. To break down this expression:
The "(?<Num>[\d\s]*)" is a named capture group that I've named 'Num' that will capture all of the numbers and whitespaces before the decimal, if there is one. This part first excludes the decimal and anything following it.
Examples:
1 2500 00 = 1 2500 00
12 500 00.00 = 12 500 00
1250 000.0000 = 1250 000
Then we use "\k<Num>" to recall that named capture group, which has the possible decimal and anything following it removed. We call up that previous capture, then scrub it for spaces using:
(\d*)\s*(\d*)\s*(\d*)\s*(\d*)
...which will remove up to 3 spaces from the value.
** CAUTION: If there are more than 3 spaces in the result, it will start to remove numbers from the end of the capture!! **
If you want to configure it to remove more than 3 spaces, simply add another "\s*(\d*)" to the end of the above.
Examples:
(\d*)\s*(\d*)\s*(\d*)\s*(\d*) - removes 0-3 spaces
(\d*)\s*(\d*)\s*(\d*)\s*(\d*)\s*(\d*) - removes 0-4 spaces
(\d*)\s*(\d*)\s*(\d*)\s*(\d*)\s*(\d*)\s*(\d*) - removes 0-5 spaces
(\d*)\s*(\d*)\s*(\d*)\s*(\d*)\s*(\d*)\s*(\d*)\s*(\d*) - removes 0-6 spaces
..... etc. .....