We study the a, b, and c coefficients of the isobaric-multiplet mass equation (IMME) using a macroscopic-microscopic approach developed by P. Möller and collaborators [At. Data Nucl. Data Tables 59, 185 (1995)ADNDAT0092-640X10.1006/adnd.1995.1002; At. Data Nucl. Data Tables 109-110, 1 (2016)ADNDAT0092-640X10.1016/j.adt.2015.10.002]. We show that already the macroscopic part of the finite-range liquid-drop model (FRLDM) describes the general trend of the a and b coefficients relatively well, while the staggering behavior of b coefficients for doublets and quartets can be understood in terms of the difference of average proton and neutron pairing energies. The sets of isobaric masses, predicted by the full macroscopic-microscopic approaches, are used to explore the general trends of IMME coefficients up to A=100. We conclude that while the agreement for a coefficients is quite satisfactory, the full approaches have less sensitivity to predict the IMME b and c coefficients in detail. The best set of theoretical b coefficients, as given by the modified macroscopic part of the FRLDM, is used to predict masses of proton-rich nuclei based on the known experimental masses of neutron-rich mirror partners, and subsequently to investigate their one- and two-proton separation energies in proton-rich nuclei up to the A=100 region. The estimated position of the proton drip line is in fair agreement with known experimental data. These masses are important for simulations of the astrophysical rp process.< Réduire