update: input data changed by OP (backets added) - consider kcott's answer : Re: regex optional word match

-*-*-*-*-

If you really-really want to use a big regex, see:

$line =~ m/^REMOTE\s+(.*?)\s+(\S+)\s+(\S+)\s+(\S+)\s+(\S+)\s+(.*?)\s+(
+.*?)sec\s+(.*)MBytes\s+(.*)Mbits\/sec(.*)/s;
    print "{$10}"; #$10 should contain at least "\n" except if you 'ch
+omp $line'
[download]

This regex become very complex if 'MBytes' could be 'Bytes', etc.

But here is another way with split (because lines seem to be well formated with "space-separator").

#!perl
use strict;
use warnings;

while (my $line = <DATA>) {
    chomp $line;
    my @parts = split /\s+/,$line;
#REMOTE Mon Jul 16 21:49:33 2012 @@ ueh1 TNT 20490 1916 0.0- 1.0 sec 0
+.33 MBytes 2.74 Mbits/sec 6.056 ms 0/ 233 (0%)
#0      1   2   3  4        5    6  7    8   9     10   11   12  13  1
+4   15     16   17        18    19 20 21  22
    next if $parts[0] ne 'REMOTE';
    
    print "IN: $line\n";
    print "\tOUT: ",join " ",@parts[1..5,8,12,13];
    print " ",  join " ",@parts[18,19] if $#parts>=19;
    print "\n";
}
__DATA__
REMOTE Mon Jul 16 21:49:33 2012 @@ ueh1 TNT 20490 1916 0.0- 1.0 sec 0.
+33 MBytes 2.74 Mbits/sec 6.056 ms 0/ 233 (0%)
REMOTE Mon Jul 16 21:49:34 2012 @@ pdn1 SSH 20499 3 1.0- 2.0 sec 0.34 
+MBytes 2.86 Mbits/sec
LOCAL Mon Jul 16 21:49:34 2012 @@ nada
[download]

Output:

IN: REMOTE Mon Jul 16 21:49:33 2012 @@ ueh1 TNT 20490 1916 0.0- 1.0 se
+c 0.33 MBytes 2.74 Mbits/sec 6.056 ms 0/ 233 (0%)
    OUT: Mon Jul 16 21:49:33 2012 TNT 1.0 sec 6.056 ms
IN: REMOTE Mon Jul 16 21:49:34 2012 @@ pdn1 SSH 20499 3 1.0- 2.0 sec 0
+.34 MBytes 2.86 Mbits/sec
    OUT: Mon Jul 16 21:49:34 2012 SSH 2.0 sec
[download]

update: /s modifier in regex