CELLULARLY RESOLVED MOLECULAR PATHWAY ASSESSMENT IN BIOPSIES VIA SPECTRAL IMAGING


Year of Award:
2009
Award Type:
R44
Project Number:
CA130026
RFA Number:
RFA-CA-07-010
Technology Track:
Molecular & Cellular Analysis Technologies
PI/Project Leader:
LEVENSON, RICHARD M.
Other PI or Project Leader:
N/A
Institution:
CAMBRIDGE RESEARCH AND INSTRUMENTATION
This is a Fast-Track application to provide reliable, cellularly resolved molecular pathway assessment in cancer biopsies to assist pharmaceutical drug development and provision of patient-specific prognosis and therapy guidance ('personalized medicine'). The organizing theme is that the appropriate unit of analysis should be the individual cell as opposed to averaged tumor extracts. To this end, novel technologies will be coupled with careful methods-development. Spectral imaging and advanced image analysis tools will permit multi-target immunohistochemical (IHC) and/or immunofluorescence (IF) detection, at the cellular and subcellular level in intact tissue sections. CRi-developed image processing and machine-learning tools provide automation and sophisticated quantitation options. Multiplexed staining protocols will yield independent, potentially stoichiometric labeling with combined IHC and IF. The sensitivity of all potential markers to variations in tissue handling will be carefully assessed; some may be robust and suitable for archival tissue studies, others will be too labile. Ultrasound-assisted fixation will be tested for its ability to preserve such labile epitopes for use in prospectively acquired tissues. Four or more pathway-related proteins will be detected in tissue sections, on a cell-by-cell basis, even if co-localized and with spectrally overlapping labels. The coordinated subcellular location of the pathway molecules will be also tracked, with simultaneous assessment of cell-surface receptors (e.g., EGFR, VEGF, Her2-neu), downstream signaling proteins and phosphoproteins (e.g., pAKT, pERK), nuclear proteins (e.g., ER, Ki67), and novel players such as protein-folding mediators (e.g. BIP1). The project will combine optimized tissue protocols, multiplexed IHC/IF reagent kits, and unique machine- learning image analysis that can be used to automate region-detection and label-quantitation. All these depend on CRi's multispectral imaging approaches for assessing multiple analytes on a cell-by-cell and cell- compartment basis in tissue sections. Our collaborators will provide small-animal tumor models for early methods development, multiplexed immunohistochemical labeling of pathway proteins in clinical cancer biopsies, access to archived and prospectively acquired tissues from pathway-targeting clinical drug trials, highly informative archival tissue microarrays, access to validated, activation-specific antibodies, ultra-fast tissue fixation, and biostatistics support. The 'deliverable' will be a suite of products suitable for clinical use that can provide much-needed valid information on single-cell-based pathway status in an intact tissue context to support pharmaceutical drug development efforts and provide molecularly focused patient care. Significance and lay narrative: The ability to quantitatively evaluate multiple molecular targets and pathways on a cell-by-cell basis, in a single preparation of clinical tissue, is missing from the current toolbox of personalized medicine, which lacks good means of matching novel drugs and drug candidates to specific patients. Conventional molecular pathology methods are typically limited to a single immunohistochemical (IHC) test on a given tissue section or to expensive and time-consuming proteomics or expression-array approaches (which cannot directly report out pathway activation status in cancer cell populations and subpopulations). Multiplexed IHC (including immuno-fluorescence) combined with optimized sample handling protocols to retain pathway proteins and advanced image analysis will enable the unambiguous detection of active signaling pathways, benefiting pharmaceutical research in the selection of patients for better targeted trials and in the monitoring of response, and clinical practice for diagnosis, therapy selection, and monitoring response (i.e., theranostics).