The message is encoded according to
RFC 2047[
^]. The encoding mechanism from your example is
Base64[
^] (indicated by the 'B').
If you don't want to use an additional library, you must write your own parser to get the code page and encoding ('B' for Base64 or 'Q' for quoted-printable) and decode the text (there are many C/C++ source examples in the net).
Once you have the decoded text, convert it to Unicode using
MultiByteToWideChar[
^] passing the corresponding code page number, and finally convert the wide string to UTF-8 using
WideCharToMultiByte[
^].