Extending "Find Usages" for existing languages

My use case is this: I've defined a custom syntax for metadata within existing language formats. It looks like this for a java file:

// <CustomSyntax Definition>
int foo;

Elsewhere in the code:

// <CustomSyntax Usage>
int bar;

(The "//" here does really mean a Java comment. That way, the Java compiler can simply ignore my custom syntax definition, while a parser I built can correctly interpret it).

I'd like to find usages of "<CustomSyntax Definition>" and find "<CustomSyntax Usage>". Is it possible to extend the "find usages" API to achieve that?

I understand this would be perfectly possible if I was defining a custom language (something like .javacustom), but I'd like to avoid that if possible.