Every coder has his/her own style, like a signature, or handwriting. I would say if you want your code to be comfortably readable by the largest audience, use the "default" styles that you generally find as the output from a disassembler, or assembly listing from a compiler. In this way copying any disassembly into your program requires no reformatting and any coder worth their coding skills will be familiar with that format, logically speaking that is.
Macros aren't generally listed in disassembly so you are on your own on that, I will say it helps to have a common symbol like @, %, or + in every macro name to allow for quick recognition of macro identifiers. Most C familiar programmers like to capitalize macro names but it isn't necessary if you affix a common symbol to your macro identifiers because assembler syntax barely resembles spoken language.
If you want to see some "established" coding style (and learn a some tricks) in X86 ASM coding, be sure to read Randal Hyde's "The Art of Assembly Programming"
I would like to give you some advice if I may be so bold. It can get pretty rough maintaining a large assembly program, try to keep things small and modular, try to write C compatible libraries so you can take full advantage of your time invested on machines where that code will execute.
People argue that assembler is not portable, well if you write the same API in C as well as assembler, you will have the option of installing the optimized version of your library on machines that can run your X86 code, and you will still have the advantage of compiling for multiple architectures.
Be forewarned this is my personal opinion based on personal experience and does not intentionally conform to any written guideline. However I am more than sure that with proper discipline, research and development you will find my method to be most rewarding.