Be careful ! The varactor is badly modelling.
The transient don't fit with the behaviour of a varactor. To check the problem, please try to charge the cap through a resistor and calculate the energie provided by the source. You don't obtain E=CxV2 (true relationship with varactor). Something wrong.
Now, try to put the integrator block outside the feedback. It works, but limited by the solver (if the order of function is greater than 1).