If you really wanted to use regex, you can use it to narrow the dictionary list before preformming one of the above (most notably Zaxo's) methods.

  my $jumble = 'RTESAMCNA';
  open FILE, "/usr/share/dict/words";
  my $n = length $jumble;
  my @possible_words = grep /^[$jumble]{$n}$/i, map {chomp; $_} <FILE>
+;
  close FILE;
[download]

This narrows (with the dictionary i have) the list for "RTESAMCNA" down to just 35 words. From there, you can use the lc/split/sort/join & hash method to get the final list (for example, it finds 'treatment', but clearly that should be excluded). Obvisouly the longer the jumble, the better off this initial pass is.. (And depends on the letter combinations, too)