my @list= qw(a b c a c);
my $w="kandabbbc";
while (my $a=shift(@list)) {
   $w=~s/\Q$a//;
   @list = grep {!/\Q$a/} @list;
}
print $w;
[download]

OMG! while and grep to iterate over @list... ok, TMTOWTDI, but then how 'bout a for loop? Also, who told you that @list has duplicates? In any case, if so, then a technique a la'

my %saw;
for (@list) {
    next if $saw{$_}++;
    # ...
}
[download]

would suffice. Also, don't you think s///g would be better suited to remove chars from his string?

Incidentally to remove charachters, the best tool would be tr///d a.k.a y///d, but unfortunately it does not interpolate - I don't know if it would be worth to use the usual eval trick to work around this limitation. The OP may still want to know just in case he really knew @list in advance and did not need to delete chars dinamically.

my @list= qw(a b c a c);
my $re = join '|', @list;
[download]

How 'bout using a charachter class instead? That is:

my $re=join "", @list;
$re="[$re]";
[download]