Figure 1

a-e. Profiler System Organization and Workflow for TMA Module. Component view of the architecture of the Profiler framework(a). The Profiler interfaces with multiple components besides TMA data. For example, the Clinical Research Information System (CRIS) component manages clinical data on patients being taking care of at the Dana Farber Cancer Institute and the Brigham and Women's Hospital. Using regulatory compliant protocols, appropriate consented cases information can be linked to the system. The karyotype component manages cytogenetic data. Different components are linked by means of unique identifiers (e.g., research coded identifier). The current system contains modules for karyotype, expression array, SNP array (not shown), and tissue microarray (TMA) data. Users can access different components using Graphical User Interfaces (GUI) at local workstations. The process of TMA use can for research can be divided into 5 steps, which include the following: research design, experiments, scanning of histospot images, histospot evaluation, and analysis. The profiler system is designed to store data for all 5 steps of this process. The first step in the work flow of a TMA study is the research design (b). As described in the text, each TMA study design is based on the questions being addressed. Three examples are given in this schematic view of research design. Test TMAs are usually small TMA with a small number (i.e., n = 100) cores. The test arrays are useful for working up reagents. These arrays provided limited information regarding protein expression or RNA expression and represent an efficient method of conserving tissues. Survey TMAs allow for determination of expression on a wide range of samples. Standard survey TMAs have approximately 500 histospots. An outcomes TMA may include a large number of samples with outcomes information. These usually represent the most valuable samples given the annotation of each case. As described in the results section, organization of the paraffin tissue blocks (upper right), circling of the areas of interest on standard slides and construction of the TMA are the final steps of developing a TMA for experiments. After the conceptual design of the TMA experiment, a physical map (TMA Map, bottom) is draw based on the available tissue blocks. The research identifiers from these blocks are linked to the research patient identifiers. Part of the experiment includes the construction of TMA (e, top). Glass slides and blocks are organized, the pathologist circles the areas of interest and tissue cores are taken from the donor block and placed into a tissue microarray. Typical types of experiments that can be performed using TMAs include immunohistochemistry (IHC), immunofluorescence (IF), in situ hybridization (ISH), and Fluorescence In Situ Hybridization (FISH) (c, bottom). Following experiments with IHC, histospot are scanned (not shown) and the data is evaluated using the Internet interface of Profiler (d). As described in the text, the system allow the reviewer to view the histospot images at several magnifications (shown is 10 ×, also available are 2.5 ×, 5 ×, and 20 × magnification). Data can be added by the reviewer (d, right side). The reviewer can also see the TMA map design (d, bottom). The growing annotated data on a TMA makes it more valuable over time. Analysis is performed on the data usually outside of the Profiler system using standard statistical analysis tools (e). Type of results that can be evaluated using TMA technology include the evaluation of individual biomarkers with respect to clinical outcomes as represented by a survival curve (e, left) or global protein expression as represented by a heat map from multiple TMA experiments (e, right).