Thanks. The problem is that we cannot reset the synchpoint as we have another applier running for the same capture.
We had a staging database in between. We had to take a 10 mins of outage on primary to reset the synchpoint. We restored the production database to the staging database to point in time when outage was started and then kicked of a full refresh to the replicate copy from the staging database. After the refresh was done, we started the applier on replicate. The applier obviously starts applying from the redefined synchpoint which was defined during the outage.