A non-orthogonal multiple access (NOMA)-inspired integrated sensing and communication (ISAC) framework is proposed, where a dual-functional base station (BS) transmits the composite unicast communication signal and sensing signal. In contrast to treating the sensing signal as the harmful interference to communication, in this work, multiple beams of the sensing signal are employed to convey multicast information following the concept of NOMA. Then, each communication user receives multiple multicast streams and one desired unicast stream, which are detected with the aid of successive interference cancellation (SIC). Based on the proposed framework, a multiple-objective optimization problem (MOOP) is formulated for designing the transmit beamforming subject to the total transmit power constraint, which characterizes the trade-off between the communication throughput and sensing beampattern accuracy. For the general multiple-user scenario, the formulated MOOP is firstly converted to a single-objective optimization problem (SOOP) via the -constraint method. Then, a double-layer block coordinate descent (BCD) algorithm is proposed by employing fractional programming (FP) and successive convex approximation (SCA) to find a high-quality suboptimal solution. For the special single-user scenario, the globally optimal solution can be obtained by transforming the MOOP to a convex quadratic semidefinite program (QSDP). Moreover, it is rigorously proved that 1) in the multiple-user scenario, the proposed NOMA-inspired ISAC framework always outperform the state-ofthe-art sensing-interference-cancellation (SenIC) ISAC frameworks by further exploiting sensing signal for delivering information; 2) in the special single-user scenario, the proposed NOMA-inspired ISAC framework achieves the same performance as the existing SenIC ISAC frameworks, which reveals that the coordination of sensing interference is not necessarily required in this case. Numerical results verify