Just because once i had a similar problem and never figured out why it hapened. How often do you run the maintenace job? As the system as high activity is it possible that when you run the maintenace job, there's a lot of "work" to be done? Could you try to run the maintenace job more often?
Also i've seen this message (warning) during a heavy process execution that produced a huge quantity o transaction log information, but when the process finishs log reader was able to catch up, in your case it seems that the replication process stops.