Assuming at least one ticket, you could do this instead:
SELECT MAX(g.ticket) FROM ticket_custom g WHERE g.ticket=id
You can convert that to a JOIN or nested SELECT etc.
so change id=c.ticket id=MAX(c.ticket)
but check your WHERE & LEFT JOIN there's some strange stuff in there.
This appears to be more of a design issue than a query issue. While it is possible to return only a single record from the ticket_custom table there is not a way to determine which record that will be. In you sample table ticket 1 has a value of "Ready for..." and "Verified" but nothing that shows which happened first. Logically, the "Ready for..." happened first.
The easiest way for you to fix this issue would be to add a timestamp or incrementing ID field. Then your subquery could select the most recent entry (latest timestamp or highest ID).
Don't make your database schema dynamic (i.e. "table-driven"). It isn't a sensible design pattern (I know from experience). It saves only a little effort upfront in creating additional attributes, but incurs additional overhead thereafter, both in writing queries and in the database processing required to execute them.
Just add a state id column to the ticket table and update it when the state changes. Create a reference table to store the state names. Then your queries become quite simple, and requires less effort for database engine to do an indexed join to a small reference table.