If I have a .docx file and add a comment, then convert it to .md, the result is something along the lines of:

[This is the comment]{.comment-start id="1" author="Mike" date="2020-12-17T16:53:00Z"} This is the content being commented on[]{.comment-end id="1"}

Here it is again with some new lines inserted to make it a bit easier to read (note it will not compile with these newlines)

[This is the comment]
  {.comment-start id="1" author="Mike" date="2020-12-17T16:53:00Z"}
This is the content being commented on
[]
  {.comment-end id="1"}

Which can then be converted back into .docx by pandoc and the file remains unchanged (the comment is still there).

However, if I add a reply to that comment (common when collaborating), the resulting .md contains:

[This is the comment]{.comment-start id="1" author="Mike" date="2020-12-17T16:53:00Z"}[Here is my reply]{.comment-start id="2" author="Mike" date="2020-12-17T17:39:00Z"} This is the content being commented on[[]{.comment-end id="2"}]{.comment-end id="1"}

Again, it's easier to read if I add newlines:

[This is the comment]
  {.comment-start id="1" author="Mike" date="2020-12-17T16:53:00Z"}
[Here is my reply]
  {.comment-start id="2" author="Mike" date="2020-12-17T17:39:00Z"}
This is the content being commented on
[
  []
    {.comment-end id="2"}
]
  {.comment-end id="1"}

So the .comment-start of the reply comment comes immediately after the .comment-start of the original comment. And the .comment-end of the reply comes in the [] brackets of the .comment-end of the original comment.

However, when I re-compile this .md, I lose that reply comment. When I convert it back to .md again, the reply comment is gone completely.

Is this deliberate behaviour from pandoc, is it an issue or am I doing something wrong here?

I have tried wrapping the comment and content in the reply comment, and all this does is add a second comment on the same content, but not a reply.