How to construct/order a TMA from the facility

WHAT IS NEEDED
To construct or have someone to construct a Tissue Microarray for you, you will need the following:

  • paraffin-embedded tissue with a non-detachable, readable label, ideally containing a barcode. The tissue must be of sufficient thickness and representativeness. For samples which are markedly uneven in thickness, there are tricks to fill this up, but it means coring twice and place one core on top of the other (re-coring).
  • an H&E stained slide or an image of it, preferably with the area to be sampled, marked. In case the sampling cannot be random, we may need to cut a superficial H&E for orientation.
  • a list (Excel spreadsheet in digital format preferred) of each and every case you need to include. Include controls. One field should contain the leading identifier of the block. Other fields may include additional data or comments you may want to be included in the TMA design sheet. If you plan to send data in a non-secure fashion, consider the use of anonymization of sensitive information.
  • a memory card, SD card, an e-mail account or any tool with which to receive the TMA design, the XML files (these are less than 1Mb files) and optionally virtual slides (hundreds of Mb each) we may produce from the TMA or donor blocks
  • instructions on the dimension of the array, which symmetry is preferred, if you need sectors and in which preferred order the samples may be laid out. We may suggest you alternatives, depending on the size of the recipient block or other considerations.
  • instructions on the number of TMAs replicas required and on whether you need tissue cores placed in vials or multiwell plates for extraction.
  • WHICH FACTORS TO CONSIDER IN DESIGNING A TMA
    These are the factors you should consider in designing a TMA: note that some of these contain mandatory items e.g. in case you need to use the TMA for diagnostics.
  • core size (0.6, 1, 2 or 3 mm) and replicated cores. This will affect the number of samples per array. We cannot mix cores of different size within the same design.
  • orderly distribution of cores/replicas in the array and presence/absence of sectors. The arrayer allows ordered or random allocation of the cores in the array (see image below)
  • See also the Excel file produced at the completion of the array (open at this link).
    Design choices
    If you choose random core allocation, which allows better quality by allowing a correction for the uneven distribution of the stain over the TMA surface, you will need to analyze the TMA digitally by exporting the design in a scanner and segmenting the virtual image.
  • source of asymmetry. For a detailed discussion on the importance of this aspect, see the JPI publication.
  • the degree of control over the identification of single cores. For diagnostic use, you should use a linear or 2D barcode throughout and ensure to have error control through all the process (see the JPI publication)
  • source of identification of each sample. The leading ID for the arrayer is what you will find in the scheme in the Excel file (circled in the example below).
  • Donor block info
    Information can be inserted in three fields in the arrayer (see image above): an ID code, a Pathology note field and a "Notes" field. Up to four consecutive fields right of the ID field in the Excel spreadsheet will be all inserted into the "Notes" arrayer field, separated by pipes (|).
    The three fields of each donor block will be reported both in the Excel sheet and in the XML files.
    EXPORTING SAMPLE DATA AND THE TMA DESIGN IN IMAGE ANALYSIS PROGRAMS
    The Galileo 4500 Tissue Arrayer provides two types of data files, to be exported in other analytical platforms. One is an Excel spreadsheet, which contains one sheet with the list of donor block IDs, and another sheet with a design of the the map and the ID of each core. It also produces two types of XML files of DICOM standard format, one compatible with the Aperio line of scanners, the other compatible with the Visiopharm and other platforms.
    For other platforms, including links with the Surgical Pathology LIS, please consult Integrated Systems Engineering Srl, scientific partner with the Facility.