the choice depends also on the throughput you need... have you got 10 patients, 100 or more?
You may decide to screen for mutations one exon at a time with methods such as DGGE or SSCP, or to PCR amplify exons (including intron-exon junctions) and sequence all of them, or focus on the exons encoding the DBD. Of course this means that most axons will turn out to be wild type. Having access to NGS may accelerate the process. But you have to think about the likelihood of finding a mutation and also about the sensitivity of each technique in identifying a mutation that may be in a minority of cells in the tumor.