In this letter, we investigate an intelligent reflecting surface (IRS) empowered robust secure transmission approach for an integrated satellite and terrestrial network (ISTN) with multiple eavesdroppers. The satellite transmits multicast signals to earth stations (ESs) within its coverage, while the base station (BS) serves many terrestrial users (TUs) with the aid of IRS. To improve the security of the satellite system, by assuming that only imperfect channel state information is available, we formulate a weighted sum rate maximization problem subject to the intercept probability constraints, the quality-of-service requirement of each ES and TU, and the transmit power budget of the satellite platform and BS. To tackle this non-convex problem, we jointly employ the Charnes-Cooper approach, the second order Taylor expansion and Bernstein-type inequality, and propose a successive convex approximation based algorithm to obtain the satellite beamforming weight vector, the BS power allocation and the IRS phase shifts. Finally, simulation results demonstrate that our proposed scheme can improve the spectrum efficiency and secrecy performance of ISTNs compared to the benchmarks.