Recommendations for Tissue Microarray Construction and Quality Assurance

Abstract

Tissue microarrays (TMAs) are important tools to conserve precious tissue resources from increasingly smaller biopsies and to control experimental costs and variation across sample sets. The quality assurance assessment of TMA materials created at centralized biobanks has not been standardized. Herein, we outline 2 processes for the construction of tissue microarrays (“recipient block” and “tape” methods) and the associated pre-construction quality control measures (pathology review, protein and RNA assessment, map creation, and storage conditions) developed by the AIDS Cancer Specimen Resource (ACSR) Network’s Science and Technology Core. These steps provide a suggested framework for quality assessment that allows end-users, receiving materials from tissue banks, confidence in their experimental results.

Introduction

Tissue microarrays originally described as “multi-tumor (sausage) tissue blocks,” are used for a variety of reasons including conservation of precious tissue resources,^{1, 2} convenience (assembly of various test/control tissue types into one block, antibody reactivity screening), reducing immunohistochemistry (IHC) costs and staining variation across different samples.^3–6 TMAs are increasingly used to study large sets of patient tissues,^{7, 8} such as those from clinical trials or other well-annotated patient cohorts.^9–14 Such tissue collections are extremely valuable and often held at centralized tissue banks.

A common method for constructing TMAs is to punch a core from a donor paraffin block using a small trocar which is then deposited into a “recipient block”. The recipient block method for TMA construction (R-TMA) was originally performed manually and later with more specific tools and automated systems.^{2, 8, 15, 16} A second manual “tape” method for TMA construction (T-TMA) has advantages over the R-TMA method as it permits working with donor blocks of variable thicknesses and allows for representative full face sections with minimal block facing. Although known to many laboratories, to our knowledge, the T-TMA method, its advantages over the R-RMA method and how to choose a given TMA construction method have not been described in the published literature.

With the increasing emphasis on rigor in experimental methods by the National Cancer Institute and others, quality assurance metrics for tissue biomarker studies are becoming more important to report in study method sections.^{17, 18} These quality measures should include documentation of desired tumor/target tissue in the core at increasing depths through the block, protein and RNA quality assessment, TMA map with orientation marker(s), and storage conditions. This technical review describes the R-TMA and T-TMA construction methods and pre-construction quality assurance steps in place at the ACSR Science and Technology Core. The decision making process involved in TMA construction for optimal use of donor blocks of variable conditions is also discussed.

Materials and Methods

Materials: The TMA kit with core tips of 1–5 mm diameters, manual punch and recipient block molds were from Quick-Ray, Sakura Finetek (Torrance, CA).^{15, 16} The 0.6 mm core punch was from Beecher Instruments (Sun Prairie, WI) .¹⁹ TMA construction consumables included: Paraffin R, tissue cassettes, paraffin molds, feather blades (all from Tissue-Tek, Sakura Finetek, Torrance, CA), and ¾ inch wide double sided Scotch tape (3M, St. Paul, MN). Immunohistochemical tissue staining was performed on 5 μm thick sections cut from formalin fixed paraffin embedded (FFPE) donor or TMA blocks using a Discovery Ultra IHC/ISH automated slide-stainer (Roche, Indianapolis, IN) and the OptiView DAB detection kits (#760–700, Roche, Indianapolis, IN) in conjunction with the following primary antibodies (Roche, Indianapolis, IN): Ki-67-clone 30–9 (#790–4286), CD20-clone L26 (#760–2531), HHV-8-clone 13B10 (#760–4260), pan-cytokeratin-clones AE1/AE3/PCK26 (#760–2595), and Blue Detection (#760–161) for in situ hybridization of U6 (#760–1210A). Hematoxylin and Eosin (H&E) stains were performed manually. Gross and detailed imaging was performed using a Samsung X4250LX photocopier (San Jose, CA) and an Aperio Digital Pathology AT Turbo Slide Scanner (Leica Biosystems, Buffalo Grove, IL) respectively. Pathology materials for all TMA’s were obtained in accordance with the ethical standards of the authors’ institutional review board (IRB) and in accordance with Declaration of Helsinki.

General Preparatory Approach

Candidate donor blocks were preliminarily screened for adequate tissue thickness. Matched H&Es were examined macroscopically to ensure they were current representations of the blocks; with a new H&E made if the block and H&E significantly differed. Microscopic review of the H&Es was performed by a pathologist to confirm diagnoses, determine the tumor content, circle tumor areas and, where applicable, non-tumor areas of interest. Gross (photocopy) and detailed (Aperio) images were recorded of the donor blocks and matched H&Es. The diameter, number of cores and the inclusion of paired normal tissue in the TMA was based on the tissue surface area, block thickness, percentage tumor, and the desired representation of tissue architecture in the completed TMA.^{9, 13, 15}

Prior to construction, a TMA map was created to unambiguously orient the TMA and all future derivative slide mounted sections (Figure 1).^{13, 16} We set an upper limit of 40 cores per TMA for greater ease in scoring and deparaffinization. Limiting core numbers also allowed for core distribution throughout the block while avoiding edge placement for better staining. Orientation and IHC control tissues were incorporated into each TMA, such as tonsil, which served as orientation and IHC controls for lymphoma.^{3–5, 7} Finally, the blocks were arranged in the work area corresponding to the map.

Figure 1.

Recipient TMA block method, designing the TMA Map; (A) The location and orientation of each donor block core in the TMA is determined prior to construction and recorded in a virtual TMA template map. (B) The punched donor block cores are then deposited into the appropriate hole in the pre-cast recipient block as per the corresponding TMA map. (C) Slide mounted sections cut from the resultant TMA block retain the pre-design orientation of the TMA map.

Recipient Block Method

Paraffin was melted at 60°C and the recipient block mold with the desired core size was warmed in a 55°C oven or on a controlled heated surface. A tissue cassette was placed on top of the heated mold (Figure 2A, 2B) and the melted paraffin was poured into the mold slowly, but continuously to prevent air bubbles from forming (Figure 2C). The filled mold was cooled at room temperature (RT) for 30 minutes before refrigerating at 4°C for 30 minutes. Once the paraffin had solidified, the mold was carefully removed from the paraffin block (Figure 2D) using steady, slow, even pressure. Tissues punched from donor and orientation blocks were inserted in the recipient block as per the TMA map (Figure 2E, 2F). Unused wells were filled with paraffin to give a uniform surface to the TMA. To securely bind and level the cores to the recipient block, the TMA was placed on top of a glass slide (paraffin side down, Figure 2G) and heated to 60°C for 10 minutes whereby (a) gentle pressure was applied to the sides of the block to bind the cores in the block and (b) the TMA was gently pressed against the glass slide to ensure the top of the cores were flush with the top of the block. The glass slide and TMA block ensemble was then transferred to an ice pack for 30 minutes to complete construction (Figure 2H, 2I).

Figure 2.

Recipient TMA block method, TMA Construction. (A) Silicon mold used to cast the recipient TMA block. (B) A cassette (shown in white) is placed atop the mold prior to the addition of the paraffin to allow the back of the cassette be submerged in the paraffin in order to facilitate strong binding of the TMA block to the cassette. (C) Melted paraffin (60°C) is then poured into the mold and allowed to cool to room temperature for 30 minutes before the mold is gently separated from block. (D) Resulting recipient block. (E) Donor blocks are then punched to collect cores from tumor rich areas indicated by H&E pathologist review. (F) The punched cores are then placed in the recipient block at positions indicated by the corresponding TMA map. (G) The TMA is then placed face down on a slide and heated (60°C) for 15 min before gently pressing the block against the slide to align cores. (H) The slide and TMA are then transferred to an ice pack to cool/solidify for 30 minutes (I) Completed R-TMA.

Tape Method

Block selection, H&E review, scanning of blocks and marked H&Es, core size and number determination parameters were the same as those used in the R-TMA method, while the map is a mirror image for the T-TMA method (Figure 3). A stainless steel block mold was used as the platform for construction of the T-TMA (Figure 4A). To assist with core placement/alignment, a paper grid was attached to the floor of the steel mold with double sided tape. An additional piece of double sided tape was applied on top of the grid to provide a surface which will hold the cores in place (Figure 4B). Using the edge of a scalpel or needle; donor block and control tissue cores were extruded and placed upright on the adhesive grid as indicated by the TMA map (Figure 4C, 4D). It is important to note that cores placed on the adhesive grid are a mirror image as T-TMAs are constructed in reverse to R-TMAs (Figure 3). Once the cores are in place, a tissue cassette was fitted to the steel mold (Figure 4E). Hot paraffin is then poured slowly and continuously into the mold (Figure 4F) after which the paraffin was cooled for 30 minutes at RT before transferring the entire mold to 4°C for 30 minutes. Once solidified, the steel mold, grid and tape were carefully removed to complete the T-TMA construction (Figure 4H).

Figure 3.

Tape TMA block method, designing the TMA Map; (A) Design of the TMA map is identical for both the R-TMA and the T-TMA blocks. (B) In contrast to the T-TMA, the cores are placed in the stainless steel mold in an arrangement that is a mirror image of the TMA map. (C) Once completed, the arrangement yields a completed TMA identical to the pre-designed TMA map. (D) Slide mounted sections cut from the T-TMA also reflect the orientation outlined by the TMA map.

Figure 4.

Tape TMA block method, TMA Construction: (A) A stainless steel block mold is employed for T-TMA construction. (B) To guide core placement, a grid is placed on the floor of the mold and fixed in place using double sided sticky tape. (C) Donor blocks are then punched to collect cores from tumor rich areas indicated by H&E pathologist review and (D) fixed upright to the double sided sticky tape at the appropriate position on the grid as indicated by the pre-designed TMA map. (E) A cassette (shown in blue) is placed atop the core-containing-mold and (F) paraffin is gently poured into the mold to submerge the pre-placed cores. (G) The mold is transferred to a frozen ice pack for 30 minutes to set before gently removing the steel mold, grid and double sided sticky tape to reveal a (H) completed T-TMA.

Results

Quality Measures

Both types of newly constructed TMAs are cut to obtain thirty serial 5 μm sections, numbering the slide mounted sections from 1–30. Pathologist H&E review is performed on sections 15 and 30, to assess core presence and tumor content. Single IHC was performed on sections 11–14 and 26–29 to assess protein and RNA quality. Pan-cytokeratin antibodies were used to detect cytoplasmic antigens. Nuclear protein quality was assessed by Ki67 staining, while mRNA quality was examined using U6 in situ hybridization.²⁰ Tumor specific IHC was also performed, such as CD20 for B-cell lymphomas or LANA for Kaposi’s sarcoma. All antibodies were scored by a pathologist and validated against appropriate control tissues. The pathology review results were documented and the stained TMA slides imaged using the Aperio Slide Scanner (Figure 5), both of which could then be provided to investigators using the ACSR TMA biobank resource.

Figure 5.

Overview of the pre and post construction quality processes; (A-C) Exemplary images of donor block H&Es with circled areas of tumor (blue) and non-tumor (green) as determined by pathologist review. (D) H&E staining, (E) Pancytokeratin immunohistochemistry and (F) U6 in-situ hybridization of a prostate cancer TMA, for which the map is shown (G) where spots 1, 2, 3, 4, 7 and 8 were derived from the circled areas in 5A, 5,B and 5C.

Discussion

TMAs are an increasingly important component of translational research, providing conservation of tissues and decreasing costs. Although the R-TMA construction method is more common than the T-TMA method, there are clear instances where the more technically demanding T-TMA method may be appropriate. Due to the pre-casting of the R-TMA recipient paraffin block, the R-TMA method works best with thick donor blocks of similar depth and will result in straighter, regularly spaced cores compared to the T-TMA method. Although it is possible to stack two or more cores from a single block on top of one another in R-TMA blocks, multiple punches from a single donor block are not always possible or well aligned. In such situations, the T-TMA method of casting the paraffin block around levelled cores, yielding cores flush with the top of the paraffin block, creates a more appealing product. Both methods have been used extensively by the ACSR Science and Technology Core to create over 30 TMAs available to investigators via the website (https://acsr.ucsf.edu). Although both methods are known in the field, little is published regarding their construction and quality control. It is hoped that this technical review may serve as resource to those involved in this endeavor.