In general antibiotic resistance is lost when it doesn't confer an advantage. In some cases resistance to antibiotics costs cells energy they'd much rather be spending on other things. When the antibiotic is gone, cells without the energy tax can outcompete ones that do and divide quicker. The population then reverts to sensitive cells again.

Even in a hypothetical case where a random enzyme mutation confers resistance with no real obvious cost or side effect to the cell, unless there is specific environmental pressure to keep that mutation, it can easily revert in subsequent generations.

Bacteria aren't resistant to a drug. They're resistant to a mechanism of action. For example, bacteria develop beta-lactamase based resistance, which can make them resistant to all beta-lactam antibiotics, which is any antibiotic with -illin in the name.

You'd potentially have to take whole classes of antibiotics out of circulation, for a long time. And even then, some of them will maintain resistance genes because these compounds are present in the environment.