Right-censored units — the ones still running when the study ends — are not missing data. They are evidence of survival.

Any analysis that quietly discards them will mis-estimate reliability.