"Design and Development of an Automated, Mass Spec-based Clinical Diagnostic System for the Detection of Ovarian Cancer."

Rick Saul(1), Ph.D., Paul Russo(2), Ph.D., Gordon Whiteley(2), Ph.D.(1)Axo Diagnostics, Gaithersburg, MD and (2)Clinical Proteomics Reference Laboratory, SAIC-Frederick, Inc., Gaithersburg, MD

In February of 2002, Petricoin et al published an article in Lancet demonstrating the utility of using proteomic patterns in the diagnosis of ovarian cancer (using Ciphergen’s SELDI technology). These findings prompted the US Congress to pass a Congressional resolution (H.Con.Res.385) in July 2002 directing the NIH to " conduct or support research on the effectiveness of the medical screening technique of using proteomic patterns in blood serum to identify ovarian cancer…" In response, the NCI contracted SAIC-Frederick, Inc. to establish the Clinical Proteomics Reference Laboratory (CPRL) with the primary goal " to develop, validate and obtain FDA approval for a proteomics pattern recognition method for the diagnosis of ovarian cancer." While many of the early studies were performed manually during the Concept Phase (Phase 1 of the GMP Design Control Process), the requirement to automate the process in order to develop a System compatible with FDA and CLIA regulations led us to explore the use of robotics.

Thus, using Design of Experiment (DOE) methodology, we have completed the Feasibility stage (Phase 2) by producing data demonstrating that the SELDI platform in combination with the Hamilton Microlab STAR is capable of meeting the specifications needed for a clinical diagnostic system. The data from these studies along with a review of the System Design will be reviewed in the presentation.