Arun,

You're right- usually the temp behaviour of caps are not modeled, in particular the varactor.
They just have the voltage gradient specified in ppm/V
In contrast, you've got the linear and quadratic tempco reported for a resistor.

What I did for a process was do a.c simulation with temp sweep and observed that tempco for mos-caps varies between +ve and -ve values.
For LC-VCO design, you're better off doing such a simulation, just with the varactor and also the whole oscillator to see how temperature's affecting it.

Cheers,
Gau