Many Phase III trials fail for lack of significant efficacy for the testing agents although their success has been demonstrated in preceding Phase II trials. One of the major reasons for this discordance is the use of different endpoints in Phase II and III trials, respectively. In oncology clinical trials, tumor response (surrogate endpoint) can be determined quickly whereas survival estimation (direct endpoint) requires a long period of follow-up. To better bridge the gap between two endpoints, we propose a novel two-stage single-arm hybrid design for Phase II trials whereby the percent of tumor size change is used as an initial screening to select potentially effective agents within a short time interval followed by a second screening stage where survival is estimated to confirm the efficacy of agents. This design can improve trial efficiency and reduce cost by early stopping the evaluation of an ineffective agent based on the low percent of tumor size change. The second survival endpoint screening will substantially increase the success rate of the follow-up Phase III trial by using the same endpoint. Simulation studies demonstrated that our novel design has improved in terms of trial efficiency, trial length, and the success rate of following Phase III trials.