I've seen earlier posts asking about this, and most responses have said that you can't pump air onto the column.
There is always an exception to every rule, because I think I have succeeded in doing just this with a Biosep S4000 column on an HP1100, pumping PBS.
Briefly, a sequence was set up to run over the weekend with plenty of buffer and a 'failsafe' to stop if the buffer dropped below 100ml. After the very first run (I must have been out of the door no more than twenty minutes), the computer crashed and the HP continued to pump for the whole weekend, ignoring the failsafe. Needless to say I am very disappointed that several days' sample pretreatments have been wasted and everything must be repeated (and I have big concerns about the computer), but I am also very concerned about the column.
The reason I think that air has got onto the column is because I purged all lines with the column disconnected and then reconnected the column - nothing appeared for about 15 minutes.
If air did get onto the column, can the column be rescued?
As I write this I am re-equilibrating the column in running buffer before carrying out a number of checks. Any suggestions that I need to do would be welcome.

Modern silica columns seem to be quite resistent to mechanical damage by gas. If your column is in very good shape (no voids, no cracks in the bed) it should be ok (assuming that no oxidation took place). Just run MP through until all gas (spikes ...?) is removed.
We use a nitrogen pressure system instead of pumps. It has occassionatly happened that it was not shut off at night, so that it ran out of MP and instead had N2 flow through it at a brisk rate, probably for several hours. Only once did it kill a column as the gas ran so long it produced fines (apparently the particles started moving and grinding each other).
This does not indicate that one should or could store columns open to air, though that also does not seem to destroy most columns very qickly. Exceptions will be materials which swell when moist or oxidize extremely rapidly.
It would be instructive to be informed about your test results.

Just to give you an update, I tried equilibrating the column, but the spikes continued for what seemed like ages. I disconnected the outlet tube from the column, left the column upright and continued to pump PBS through the column (at a low flow rate). Very fine bubbles were coming out of the column for about one and a half hours, and eventually the outflow settled down (even increasing the flow rate failed to release any more bubbles). Reconnecting the column to the detector revealed a very good and steady baseline, with no sign of spikes (the detector and other components were cleared of air with the column out of line). I've decided to leave the column equilibrating overnight (with plenty of buffer!) and will test the column tomorrow. The signs are looking good at the moment.

Thatīs to be expected, it took a long time to get the gas out even at normal or higher flow rates, especially in an old column. The position of the column during flushing was irrelevant.

Finally got round to testing some samples on the column. No problems seen, retention times are unchanged, peak areas okay, etc. In fact, it looks so good perhaps I should run the column dry on a regular basis!

Its nice to have a confirmation on such an observation. The good performance of the column might be due to the extensive washing.