I think your result is not correct, Mahmoud, sorry!
It neglects the relatively low input impedance at the source of this common-gate stage, which is in parallel to R4 and might be comparably lower than this, and the output resistance Ro of M1, as well as the parallel (to RD) load resistors R2 and R3, which are connected to a virtual GND.
Your gain result is in accordance with that one of a common-source stage (at least to 1st order), and - by this - much too large for a common-gate stage.
At least I think so. I didn't retrace the solution manual solution, however.