I think I might be missing something here.........In the absence of any pulished probe/primer sequences I presume you've obtained a refseq mRNA and designed your own set of probe/primers for each target gene?  If so what exactly is it you're BLASTing these resultant sequences against?  You only need to be sure they're designed across intron/exon boundaries.

@Marcus Hanley

I can not find the mRNA sequence of HER2, BCR/ABL and Src, only transcript variants and since I do not know what part my drug targets I would need the whole sequence. Do you have any suggestion what I could do? I cannot design a primer without a mRNA sequence

I don't know where to obtain my refseq.

Firstly you'll find everything you need on the NCBI website.  Secondly I think you need to understand a little more the results you are hoping to gain.  If you're simply looking at effects of a drug on the expression of certain genes you don't need to know where the drug binds, you just need to know whether there is more or less of that RNA transcript in your sample and as such a probe/primer set designed against any part of that gene will be adequate.  Genes can have many names so using the 'gene' database on NCBI find out the alternative names as the RNA sequences may be listed under other names than the ones you've listed.  Performing a quick search I found Human HER2 and can see that this is a miRNA gene.  As it is very short there won't be (I've not BLASTed against genomic so I don't know for sure) any Intron/exon boundaries so you need to be sure you don't have any contaminating genomic DNA in your samples as the probes/primers will not discriminate between cDNA derived from RNA transcripts and genomic DNA.