The T2K (Tokai-to-Kamioka) experiment is a second generation long baseline neutrino oscillation experiment that probes physics beyond the Standard Model. An off-axis neutrino beam with a peak energy of $\sim0.6$ GeV is produced at the J-PARC accelerator facility, with the flavour content dominated by either muon neutrinos or muon anti-neutrinos, depending on the choice of the polarity of the magnetic focusing horns. The neutrino beam is detected first in the near detector ND280, where the flavour composition of the incoming neutrino flux is not expected to be affected by oscillation, and then travels 295 km to the far detector Super-Kamiokande, where oscillation significantly affects the flavour composition. We report the results of a joint analysis of neutrino and antineutrino oscillations at T2K with the $\nu_\mu$ ($\bar\nu_\mu$) disappearance and $\nu_\mu$ ($\bar\nu_\mu$) $\rightarrow$ $\nu_e$ ($\bar\nu_e$) appearance channels, obtained by collecting a total statistic of $7.57\times10^{20}$ protons-on-target in $\nu$-mode and $7.53\times10^{20}$ in $\bar\nu$-mode. The results in the disappearance channel shows that our data continue to prefer maximal $\theta_{23}$ mixing ($\sin^2\theta_{23}=0.532_{-0.068}^{+0.046}$, in case of normal mass hierarchy) and no violation of the CPT theorem, while in the appearance channel, we observed a large $\nu_e$ appearance and a low $\bar\nu_e$ appearance with respect to the expectations. These results favour a $\delta_{CP}\sim-\pi/2$, with a 90% confidence interval of $[-3.13,-0.39]$ in normal mass hierarchy and $[-2.09, -0.74]$ in inverted mass hierarchy. We comment briefly on the future prospects for T2K, including a proposal for extended running to accumulate $20\times10^{21}$ protons-on-target, nearly three times the currently approved amount by 2026, to gain substantial sensitivity to CP violating effects in $\nu_\mu$ ($\bar\nu_\mu$) $\rightarrow$ $\nu_e$ ($\bar\nu_e$) oscillations if parameters are favorable.