• Home
  • News
  • Calendar
  • About DF/HCC
  • Membership
  • Visitor Center

Tissue Microarray and Imaging Core

Fig. 1. Example of colon cancer tissue sample double stained for pancytokeratin (A) and TS (B). Nuclei are stained by DAPI (C). The merge image is shown in (D).

The Tissue Microarray and Imaging Core is dedicated to the construction and evaluation of high quality tissue microarrays for cancer research. Core services include construction of standard and custom tissue microarrays, preparation of slides, image acquisition and analysis as well as consultation and high throughput isolation of DNA and RNA from formalin-fixed paraffin-embedded tissue cores.  

The core has introduced new services including:

1. Proactive generation of test TMAs
The core is actively retrieving archival cases of common cancers using DF/HCC's Pathology Specimen Locator Core. These cases are used to prospectively build TMA and nucleic acid test cohorts, which will be available to users who wish to screen particular tumor types for certain biomarkers. It is anticipated that most of this work will be completed before the end of this year.

2. Reconstitution of TMA imaging function
The core recently purchased the Aperio scope and image analysis software. Using digitized slide images accessed through a web-based digital pathology information management system (Spectrum), users will be able to remotely visualize and score biomarker staining results, either directly or with the assistance of pathologists (see images below).

3. Quantitative measurement of biomarkers
AQUA and Aperio will allow the core to quantify biomarkers detected with fluorescent and immunohistochemical stains, respectively.

Recent core projects

The core worked with Susan Hankinson, MPH, ScD (HSPH) on a project validating the use of tissue microarray technology in ovarian cancer. Tissue microarrays (TMAs) allow high-throughput evaluation of protein expression from archived tissue samples. In this study, the authors identified and evaluated characteristics specific to ovarian cancer that may influence TMA interpretation. TMAs were constructed using triplicate core samples from 174 epithelial ovarian cancers. Stains for p53, Ki-67, estrogen receptor-{alpha}, progesterone receptor, Her-2, WT-1, cytokeratin 7, and cytokeratin 20 were evaluated by intraclass correlation coefficients, Spearman correlation coefficients, the effect of sample age, and tumor histology on the ability to score the cores, and inter-rater reliability. The interclass correlation coefficient and the mean Spearman correlation coefficients among three cores were ≥0.91 and 0.87, respectively. Tissue age and tumor histology were not predictive of an inability to evaluate stains, but borderline tumors had a 2 to 4-fold increase in the risk of having uninterpretable cores over invasive tumors. There was moderate to substantial concordance between the two pathologists for estrogen receptor-{alpha} [Cohen's Kappa ({kappa}), 0.79] and Ki-67 ({kappa}, 0.52). The prevalence of positive staining cells by histologic type was comparable with previous studies. In conclusion, results from this study indicate that TMA is a valid method for evaluating antigen expression in invasive ovarian cancer but should be used with caution for borderline tumors. The authors suggest several methods of quality control based on intercore comparisons and show that some antigens may be affected by age of the samples.

The core also worked with Monica Bertagnolli, MD (BWH) conducting AQUA Analysis of Thymidylate Synthase in colorectal carcinoma. There is evidence that increased thymidylate synthase (TS) expression is a marker for decreased survival in colorectal cancer. Specifically, it has recently been suggested that the relationship of nuclear to cytoplasmic thymidylate synthase expression is a strong predictor of colon cancer survival.

To test this hypothesis, the prognostic value of TS expression as a function of subcellular localization is being assessed by AQUA in TMAs containing tumor tissue cores from 936 patients with stage III colorectal cancer. TMA tissues sections were first stained with an antibody against pancytokeratin (rabbit polyclonal, 1:100 dilution; DAKO, Carpinteria, Calif) and visualized with a green fluorophore (Alexa 488) to differentiate tumor from surrounding stroma and/or leukocytes (cytoplasmic mask). Sections were then stained with the antibody to TS (mouse monoclonal clone TS106, 1:100 dilution; LabVision NeoMarkers, Fremont, Calif) and visualized with a Cyanine-5 fluorophore (red). The nuclear compartment (nuclear mask) was defined by staining with DAPI (blue). Multiple, monochromatic, high-resolution (2048 × 2048 pixels) images were then obtained (Fig. 1). For each tissue core, the quantification of the TS protein in the nuclear or in the cytoplasmic compartment was performed by generating an AQUA score corresponding to the ratio between the sum of pixel intensities for the target, and the area of the compartment of interest (mask). The TS nuclear, cytoplasmic, nuclear to cytoplasmic ratio scores are now in the process of being correlated with clinical outcome.

In this project, the core performed double immunofluorescence staining of TS and pancytokeratin in TMA sections and semiautomated quantification of TS expression in the nuclear and cytoplasmic tumor cell compartments using AQUA.

To access the core's services, please contact Tyler Caron at 617-525-7908 or tcaron2@partners.org.

Examples of enhanced imaging using Aperio technology

Automated Quantification of Immunohistochemical Staining in TMAs

Image A. Section from a TMA containing renal cell carcinoma tissue cores was immunostained for HIF1alpha.

Image B. The Aperio Nuclear Algorithm was utilized to quantify HIF1 alpha expression in the nuclei of the tumor cells. The algorithm generates a “mark up image” in which negative nuclei are labeled in blue, 1+ nuclei are labeled in yellow, 2+ nuclei are labeled in orange and 3+ nuclei are labeled in red.  Percentages of negative, 1+, 2+, 3+ nuclei as well as other image analysis statistics are calculated by the system.


Assessment of Protein Co-localization by Image Analysis of Double-immunostained Slides

The composite figure below shows p63 and fatty acid synthase (FASN) co-localization in the mouse urogenital sinus (UGS) epithelium as analyzed using the Aperio system.

Image a. Coronal section of the UGS from a wild-type E18.5 embryo double-immunostained for p63 (Fast Red) and FASN (DAB) and counterstained with hematoxylin.
Image b. Deconvoluted image of the Fast Red color channel.
Image c. Deconvoluted image of the DAB color channel.
Image d. Processed pseudocolorized image showing the DAB channel (FASN) in green, the Fast Red channel (p63) in red, and the hematoxylin channel (nuclei) in blue. Expression of both FASN and p63 proteins is observed in the vast majority of cells (black arrow). Cells lining the UGS lumen (arrowheads) are negative for both markers (modified from Sabbisetti et al, PLOS One 2009).