your running a solvent program with DCM --> methanol or methanol --> DCM conc gradient?
Anyway... It would depend on which of the solvents the analyte is more soluble in, but in HPLC material does not usually get 'stuck' on the column, if it is being retained and the peak is broad this usually means your flow rate is too low and diffusion is occuring. And if you are concerned about the higher concs being retained (where or how i dont know) then lower the concs of all standards and unknowns by ratio to eliminate this problem, you should not be using such concentrations which would flood the detector (possibly damaging it) in any HPLC analysis.
However you can also just do a few blank runs and run the complete sequence for the time of analysis and while running the temp program between actual runs (this would waste alot of resources and instrument time though you realise?)
-Josh