Hi Mahdi,
In Monte Carlo simulation, you have some factors (variables) that can take values following a given probability distribution. When you run a MC simulation, you sample different values from the distribution and see (through computer simulations) how the system behaves, observing/measuring the resulting output produced. Since you are dealing with random variables, you need to run the simulations many times, and from the results you get you are able to build confidence intervals associated to the output.
Sensitivity analysis helps you establish how sensitive the system is to a given input (or combination of inputs), determining the correlation between the input and the output. Here you measure how a variation in the input(s) impacts on the resulting output.
Hope this helps,
Gabriella