Agree with Tim. The chance of finding a mutation with this degree of response is very unlikely and there is no role in my mind for mutation screening on a routine basis using standard Sanger sequencing (15-20% sensitivity). A mutation search should be triggered by a loss of response, usually a loss of MMR before one will be picked up. High resolution mutation testing (<1% sensitivity), as a predictor for future mutation development is an exploratory test, which should not be deemed as standard.
In regard to pushing the dose back to 100mg. This might not be necessary if the patient retains response to 70mg. There are recent data from the French OPTIM study which suggests that based on dasatinib blood levels, the optimal dose of dasatinib may be around 70-80mg. I would restart when you have the cleared the effusion and monitor, increasing the dose only if there appears to be loss of response.
Definitely feel that with one effusion and no loss of response with dose modification, there is no need to switch TKIs.