In clinical trials, multiplicity adjustment problems arise in claiming treatment benefit in a subgroup of patients, such as a biomarker positive population which is more likely to benefit from the drug, in addition to the overall study population. Song and Chi (2007) proposed a two-stage closed testing procedure for testing an endpoint in both populations with optimal power while strongly controlling for overall familywise error rate (FWER). In practice, besides benefit as exemplified in primary endpoint in both populations, secondary endpoints of interest often need to be tested in both populations as well. We extended previous works to allow for testing multiple endpoints in both overall and subgroup populations. The conditions to control for the FWER under different scenarios will be discussed based on simulation studies and further illustrated by real world trial examples.