Many of innovative biologic cancer drugs were approved based on late endpoints (PFS or OS). For biosimilar products currently being developed to these biologics, early endpoints (including PD biomarkers) may be used for designing biosimilar clinical trials in sensitive disease models. Such trials may serve as last step confirming biosimilarity and allow for extrapolation to other indications based on the step-wise and totality of evidence approach. To successfully optimize and implement such a biosimilar program, one of the key components is to identify an early endpoint or PD biomarker that is a surrogate or explains biological response of the agent. Such a surrogate may be based on an extensive overview of trials that give reliable estimates of the net effects of the intervention on biological response or a clinically meaningful endpoint. We illustrate through an example how to use information from historical trials with the innovative biologic based on causality analyses as suggested by several authors (Buyse et al (2007), Burzykoski et al (2001)) and recently by Fleming et al (2012). The surrogacy or semi-surrogacy will be explored by the estimation of correlation coefficient between treatment effects on the early and late endpoints using liner regression models with and without measurement errors. Justifying such a semi-surrogate biomarker requires cross-functional collaboration to bridge various expertise and activities. We explore possible organizational structure to support such translational research in biosimilars.