Laboratory of Molecular Design
Early Detection of Pancreas Cancer Onset by PET Imaging of KRAS and HER2 mRNA


     Many receptor-specific ligand candidates predicted in silico fail to demonstrate their calculated affinity and specificity after synthesis and testing. We hypothesize that docking of three-dimensional (3D) projections of potential target binding agents with 3D projections of macromolecular disease targets visualized on a computer with touch and feel feedback will enable identification of optimal agent designs, and culling of suboptimal agent designs, prior to synthesis. The use of haptic feedback and quantitative measurements of obstacles encountered along the kinetic pathway will enable culling of unfavorable designs, allowing us to synthesize more sensitive and specific molecular probes to support the projection of 3D gene product images of patients superimposed on 3D anatomical images of patients. We will test this hypothesis by synthesizing a genetic probe carrying a newly designed ligand and measuring its actual affinity for the target receptors on human pancreatic cancer cells.
     Anatomical details visible by X-rays, ultrasound (US), computerized tomography (CT), or magnetic resonance imaging (MRI) do not necessarily delineate the sites of the most serious trauma or disease. We assert that the addition of gene product imaging agents for proteins and nucleic acids by positron emission tomography (PET), SPECT, or MRI will delineate the sites of the most serious trauma or disease, because those are the places where the disease genes are most active. We hypothesize that 3D gene product images of patients superimposed on 3D anatomical images of patients visualized on a computer with touch and feel (haptic) feedback will enable improved preoperative planning, because gene product images will reveal specific sites of intense disease gene expression. This approach will provide tactile and visual feedback to the surgeon planning a procedure, just before entering the operating room. We will test this hypothesis by evaluating the perceptions of experienced surgeons attempting to identify malignant pancreatic lesions via haptic manipulation of 3D simulations of diseased human pancreas CT scans, with or without the merged FDG PET data.
     We propose to test our hypothesis through the execution of two interrelated aims that move in synergy toward the future goal of genetic image-guided surgery.