Liquid biopsy vs tissue biopsy for T790M mutation - 1272365

I recently saw the results of a German study of 1100 patients that indicated that the sensitivity of liquid bx for T790M testing for NSCLC patients was approximately 50%, compared with the gold standard of tissue bx. If this is accurate, would it make sense to do a liquid bx for the mutation and then proceed to tissue bx if the liquid bx result is negative? This strategy might avoid an unnecessary tissue biopsy I ask this question in anticipation of the need for Tagrisso ( or the analogous Clovis agent), my wife having developed resistance to Tarvceva after an excellent result for 15 months.
Thanks in advance