The Biomarker Imaging Research Laboratory (BIRL) is a cutting-edge research facility that develops new histopathology techniques and image-processing tools to enable imaging pathology correlation analyses in cancer and other diseases. It is a fundamental resource within the Centre for Research in Image-Guided Therapeutics and a core Lab that provides services to internal and external academic and industry customers.
A key focus of the lab is biomarker panel development, and validation and application through the use of multiplexing technologies. Our immuno-fluorescence multiplexing (MxIF) digital systems provide cellular quantitative information that help characterize cancer and its microenvironment for precision therapies and better disease management.
There are currently two multiplexing systems in the laboratory: CellDIVE (Leica) and PhenoCycler Fusion (Akoya Biosciences).
We support imaging studies that combine MxIF, molecular profiling and imaging by developing image-processing algorithms for digital pathology and in vivo medical imaging.
One of BIRL’s unique skills is the production of large, whole organ diagnostic slides through an alternative tissue-processing method called “whole mount”. Over the years the lab developed a full range of digital whole-mount histopathology techniques that produce large (up to 5” x 7”) tissue slides from breast, prostate, tongue, colon, pancreas and brain cancer, brain and other tissue samples.
The lab conducts its own and offers support for clinical studies and clinical trials. It operates under the licence and supervision of the department of anatomic pathology of the Sunnybrook Health Sciences Centre.
It has implemented clinical standards for quality control and process monitoring, including clinically compatible tissue tracking and sample storage systems (STS). Several state-of-the-art digitizers are available to accommodate various sample sizes, and to perform bright field and fluorescence scanning of regular and large whole-mount slides.
In collaboration with other labs, our work is focused on quantitative 3-D histopathology, including automated image processing and computer detection of structures. As a core lab, BIRL is also providing pathology services, both regular and whole mount, for preclinical research studies for imaging interventions in cancer, cardiac, neuro, regenerative medicine and other areas of research.
We offer a wide range of immunohistochemistry services for clinical and preclinical research projects and clinical trials. These services are available to internal and external investigators, clinicians, and industry collaborators and investigators. We are committed to exploring new ideas and technologies and welcome new collaborations.
The lab supports collaborations and offers services in these priority areas:
- biomarker panel development and validation using multiplexing immunofluorescence
- whole-mount tissue processing and large format tissue slide preparation (up to 5”x7”)
- advanced and regular immunohistochemistry for whole-mount and conventional tissue sections
- multimodality image processing and algorithm development, cross-platform image registration and validation
- large capacity, regular and large size tissue slide scanning in bright field and fluorescence modes
- advanced pilot projects and full studies to conduct genomics and transcriptomics profiling research through links with the genomics core facility.
Examples of services:
- Biomarker multiplexing immunofluorescence application and analysis on formalin-fixed paraffin embedded tissue samples and tissue microarrays
- tissue processing and immunohistochemistry, for regular and whole-mount tissue sections (slide sizes 3″ x 4″ and 5″ x 7″)
- ultra-rapid tissue-processing protocols for large tissue specimens
- quantitative histological examination of human and preclinical specimens
- antibody staining and optimization
- frozen sections (1″ x 3″)
- microscopy and digitization of tissue slides (1″ x 3″ to 5″ by 7″) in bright field and fluorescence modes
- microcomputed tomography scanning and image analysis
- analysis algorithms, including 2-D and 3-D visualization and correlation tools
- pathological analysis and annotations, immunohistochemistry scoring.
To place an order, download the service request form (Fillable PDF Form).
The support for the lab comes from collaborative research projects funded by grants. In addition, we provide core services using a cost-recovery approach. Below are examples of fees we charge. Since each project is unique, a detailed estimate will be made for your project. For a customized estimate, fill in and email the order form (PDF) to yulia.yerofeyeva@sri.utoronto.ca; or call 416-480-4265 or 416-480-6100 ext. 67479 to arrange a meeting.
Please note that there is an SRI-regulated overhead charge for external academic and industry collaborators. A material transfer agreement or a collaboration agreement might be required.
Tissue processing and embedding
Large tissue specimens from lumpectomy, mastectomy, prostatectomy, oral cancer surgeries are processed and embedded in our laboratory. These services are available for a fee. Please contact the laboratory manager for a price estimate.
Freezers/tissue banking
The biomarker imaging research lab is equipped with monitored -80C and -20C freezers and fridges.
Immunohistochemical and special staining
Examples of special stains routinely made in the biomarker imaging research lab:
Masson trichrome stain, Prussian blue stain, Picrosirius red stain, Alizarin red stain, Goldner’s trichrome stain, Luxol fast blue stain, Bielschowsky stain, Pentachrome stain, Verhoeff-Van Gieson (VVG), Pimonidazole (hypoxia) stain.
Examples of immune stains used and/or optimized in the lab:
14-33 (Santa cruz), ALDH1 (BD Transduction Laboratories), CD105 (Millipore), CD206 (Abcam), CD24 (SN3b) (Thermo), CD31 (Dako), CD34 (Dako), CD44 (156-3C11) (Thermo), CD45 (Dako), CD68 (Abcam, Dako), CK5/6 (Millipore), Cox2 (Invitrogen), Cytokeratin (AE1/AE3) (Dako), D240 (Abcam), ER (6F11) (Novocastra), GFAP (various), (Her2/Neu (SP3) (Thermo), Ki67 (Thermo), Neurofilament (Millipore), Osteopontin (Novus Biological), p16 (discontinued) (Millipore), p21 (Various), p53 (Various), PR (16) (Novocastra), RAM 11 (Dako), SMA (Sigma).
There are many other antibodies that we have optimized in our lab. However, each tissue type and a new antibody clone might require optimization on tissue of interest and control samples. A one-time new antibody optimization fee, if required, is about $600 for internal customers.
Examples of fees charged per slide for immunostaining are based on 10 slides processed in one batch. The fee varies greatly depending on order size, slide size and the antibody, detection or dye used. The ranges below are estimates only. A one-time new antibody optimization fee, if required, is about $600 for internal customers.
| Slide size | 1″x3″ | 2″x3″ | 3″x4” | 5″x7″ |
|---|---|---|---|---|
| IHC staining cost/per slide | $30–$65 | $80–$220 | $140–$375 | $180–$550 |
Scanners
TissueScope LE LE120 scanners
The advantage of the Tissuescope LE scanner is the ability to quickly digitize whole-mount large (up to 5”x7”) slides as well as regular 1”x3” slides in bright-field at 50X (0.2 μm/pixel) maximum resolution. Please contact the lab for a specific quote and calendar booking.
This instrument is available for self-service with proper training. The hourly rate for use of the instrument is $50 for internal users to offset the costs of maintenance and repairs.
| Slide size/magnification | 1”x3” | 2”x3” | 3″x4” | 5″x7″ | ||||
| 10X | 20X | 40X | 10X | 20X | 40X | 20X | 20X | |
| Internal rate (per slide) | $5 | $7 | $9 | $10 | $15 | $20 | $25 | $50 |
|---|---|---|---|---|---|---|---|---|
Olympus VS120
Olympus VS120 system allows for high-resolution fluorescent scanning of slides (up to X100 and over optical resolution). The hourly rate is $50 for internal users.
To inquire about space availability and fees please contact Yulia Yerofeyeva at 416-480-4265 or yulia.yerofeyeva@sri.utoronto.ca.
The biomarker imaging research lab is equipped with state-of-the-art research pathology and microscopy instruments. Many of them are available for use by other labs.
Multiplexing immunofluorescence imaging system
Cell DIVE (Leica Micrsosystems) and Phenocycle Fusion (Akoya Biosciences) allows studying co-localization of multiple biomarkers on the same tissue slide.
Tissue processing and staining equipment
These instruments allow fixing and processing of tissue samples by applying conventional and rapid automated microwave protocols, tissue embedding in wax and sectioning of various sizes (from 1”x3” to 5”x7”).
- Pathos automated microwave, Milestone
- Microwave Histos 5 tissue processor, Milestone
- ATP tissue processor, TBS
- Automated immunostainer, Dako
- Microtome SN2500/ultramiller 2600, Leica
- H&E Varistainer, slide agitator
- Automatic immunohistochemistry (IHC) stainer
- Robotic large slide staining system (designed in-house)
- Ventana Discovery instrument for IHC/in situ hybridization provides highest level of protocol flexibility for the most challenging assays and for rapid assay development
- Specimen storage facility and tracking system (for storing slides and tissue blocks)
Microscopes and tissue slide digitization and scanners
These microscopes are used to examine and digitize variously sized tissue slides in fluorescent and bright field modes at high resolutions (x10, x20, x40 and above). Training can be provided to use these instruments.
- TissueScope LE
- TissueScope LE120
- Olympus VS120
- Zeiss Axiozoom A.1
- Akoya PhenoImager
Software and image analysis tools
We operate tissue slide viewing and image analysis software using commercial and open-source viewers. We also have file converters allowing users to view images on various platforms.
- Various images viewers: Sedeen Viewer, Huron viewer, QuPath
- ImageJ
- HALO
- Merkator
- Olympus software
- ClearCanvas, etc.
- Lab-developed algorithms, including titling for large images, vessel counting, ImageJ plug-ins, etc.
MicroCT system
MicroCT specimen imaging system allows CT imaging, volume reconstruction of small samples for the purpose of imaging-pathology correlation. System resolution is 100 um.
Genomics facility
The genomics facility that is part of the biomarker imaging research lab offers in-depth sequencing and many types of molecular analysis. Read more information about the genomics core facility.
Projects
The biomarker imaging research laboratory aims to develop new quantitative imaging pathology techniques through collaborations in the following areas:
The biomarker imaging research lab specializes in whole-mount histopathology, which captures an entire cross-section of a surgical tissue specimen. These tissue sections are up to 120mm x 170mm in size and, once stained and scanned, can be used in imaging-pathology correlative studies.
Whole-mount histopathology provides improved sampling through preserving spatial arrangement of the tissue, which is key for studying disease characterization and progression such as spatial heterogeneity and progression of cancer. Other advantages of this technique are:
- more accurate surgical margin assessment;
- volumetric disease burden; and
- secondary disease foci identification
Over the years the lab has processed a large variety of whole mount tissue samples: breast cancer (lumpectomies and mastectomies), whole mount prostate cancer, oral (tongue) cancer, colon cancer, and lately, pancreas, brain and various arteries.
These large and regular tissue slides can be digitized in our lab using large field-of-view microscopes and tissue scanners such as Tissuescope LE. Tissuescope can produce high-resolution (25X and 50X) digital images of the slides, which can be converted into 2D and 3D datasets.
We are continuing our work on the improvement of the whole-mount process by optimizing laboratory procedures, using advanced equipment and applying automation to the workflow.
We have developed a prototype automated stainer for whole-mount slides to reduce further the cost and time associated with the method, and validated it on several clinical whole-mount hematoxylin and eosin slides.
Examples of automation include prototype automated stainer, automated microwave processing methods that significantly reduces overall processing time.
A Tissue slicer was developed and patented for improved sampling of large tissues and has been used routinely to render accurate slicing of the specimens to produce serial sections at equal intervals.
To accurately correlate pathology and in vivo imaging, we have developed an alternative gel formulation that allows rapid embedding and accurate slicing and sampling of the tissue. providing a multitude of opportunities for radiomic studies.
A major focus of the lab is the development and validation of reliable quantitative prognostic and predictive biomarkers.
Two multiplexing technology platforms are being used and available for all researcher and clinical collaborators: Cell DIVE (Leica Micrsosystems) and Phenocycle Fusion (Akoya Biosciences)
While CellDIVE platform (Figure 5a) uses cyclic immunofluorescence with compatible commercially available antibodies, the Akoya’s platform (Figures 5 b) uses a similar approach but with proprietary barcoding technology and custom antibody panels.
Both systems, while using different approaches, allow for co-localization of up to 100+ biomarkers on a single cell providing exciting possibilities for research on treatment outcome and precision medicine.
Biomarker multiplexing systems
CellDIVE imager (Leica Biosystems):
- Antibody agnostic stain-image-bleach cyclic immunofluorecense system
- Step 1: Antigen retrieval, serum blocking
- Step 2: Label section with CV 3, 3.5, 5, 7 conjugated antibodies
- Step 3: Acquire IF image
- Step 4: Photo-induced chemical bleaching
- Step 5: Acquire background image, repeat
PhenoCycler-Fusion System (Akoya):
- Step 1: Slide preparation and panel assembly
- Slide preparation using standard antigen retrieval and serum blocking protocols
- Step 2: Application of oligonucleotide barcoded antibodies
- Each antibody is tagged with a unique oligonucleotide barcode. Antibodies and reagents are proprietary and provided by Akoya in batches of 10 applications
- Step 3: Iterative cycling and imaging
- The platform works by staining the tissue is a single-step and iteratively cycling and imaging fluorescently labeled antibodies across the tissue through a reveal-image-remove cycle
We are involved in a number of collaborative projects developing and testing biomarker panels for breast, lung, and lymphoma cancers. We have collaborated in the past on ductal carcinoma in situ (DCIS), immunotherapies for ovarian cancer, immune checkpoint inhibitors in breast cancer and immunotherapies in hypermutant cancers including brain and lung cancers.
Highlights of current projects include a collaboration with the OICR Adaptive Oncology program, where we have studied a collection of whole-mount processed breast cancer specimens using biomarker multiplexing and found large levels of intra-tumoural heterogeneity. There we identified various clusters of cancer-associated fibroblasts, and measured if certain groups are more prevalent in breast cancer subtypes (Figure 6).
Using quantitative methods, developed by our team and commercial platforms that measure biomarker signal we are able to define the extent of cancer heterogeneity by characterizing each clone, and investigate potential functional relationships between different biomarkers. This information can be used to delineate further subtypes of cancers for developing treatments that can target heterogeneous tumours.
Our lab have been focusing on developing quantitative image processing and image analysis tools for multiplexing immunofluorescence (MxIF). Several open source software tools have been developed:
- DeepCSeg plug-in for CellProfiler
- Semi-automated radiomics pipeline for the correlation of 3-D in vivo imaging to histopathology
- Graph-based, K-nearest neighbor algorithm to quantify spatial features in cell types defined with MxIF data
- AI models developed in the Martel lab
- cellularity assessment,
- mitotic cell detection and
- TIL quantification in breast cancer
- Various display and analysis tools such as cBioPortal
We develop and use databases for linking our research and clinical data:
- SRI Biomatrix
- Specimen Tracking System (STS)
Our lab also uses software for data analysis developed by other companies:
- PathCore Flow
- HALO
Using many of these tools, we have completed a radiomic study co-funded by the DoD on the multi-omic investigation of screen-detected breast cancer. In this study we have identified RNA expression signatures of sub-groups of Luminal A cancers that showed lower risk (LR) vs higher risk (HR) profiles and a radiomic signature that correlates with the risk of recurrence provided by the 95-gene assay (Bayani et al, 2017).
The goal of this project is to find molecular changes that are linked to the imaging/radiological features, which may provide a better understanding of the phenotypic changes as a consequence of the underlying biological alterations, toward developing better tools for planning therapy and monitoring treatment response.
In a collaborative study with GE Global Research we conducted a radiomic analysis of high risk breast cancer cohort of women by analysing their dynamic contrast-enhanced MRI image sets and correlating them with the genomics and transcriptomics data from their histopathology specimens. We were able to associate genes or pathways related to features computed from images, and then identify their potential correlations to clinical outcomes.
In a multi-omic study of breast cancer funded by the US Army Breast Cancer Program (2021-23) we used samples collected from the TMIST Lead-In trial analyses radiomic, molecular and proteomic signatures. Early data from a small sample size allowed us to investigate potential correlation between MRI radiomic features to RNA to protein expression clusters, the radiomic features to the 95-gene recurrence risk score. (Figure 8a)
Fig. 8a: Correlative radiomics in a breast cancer heterogeneity study
In an ongoing prostate cancer heterogeneity study, “Radiomic mapping of heterogeneity in early prostate cancers – targeted radiomic imaging for heterogeneous cancers”, we seek to train MRI features linked to a biomarker signature for occult high-grade prostate cancer (PCa), called PRONTO. Here, we construct radical prostatectomies (RP) from whole mount and sectioned prostates from H&E-stained slides into both 2D and 3D constructs, enabling registration of tissue sections and respective tissue blocks to the corresponding MR slice. (Figure 8b)
PathcoreFlow:
Fig. 8b: Correlative radiomics in a collaborative prostate cancer heterogeneity study.
The Biomarker Imaging Research Lab is a state-of-the-art facility dedicated to early-phase imaging clinical trials and pilot studies. It has been integral to numerous clinical trials validating innovative imaging techniques and biological markers for disease characterization.
BIRL is involved in a number of clinical trials:
- Tomosynthesis Mammographic Imaging Screening Trial (TMIST)
- Whole-mount breast cancer method (whole mount mastectomies and lumpectomies)
- Photoacoustic imaging of prostate cancer in ex vivo post prostatectomy specimens
- Correlation of [18F]fluoroazomycin-arabinoside (FAZA) PET hypoxia imaging to 3-D histology in oral tongue cancer (PI: Ian Poon)
- Molecular and imaging biomarkers for precision therapy of breast cancer (recently completed).
BIRL Lab is accredited to conduct clinical studies, with a specialized focus on whole mount breast, prostate, and oral cancer research. We perform semi-annual and annual preventive maintenance to ensure optimal equipment performance. Additionally, we utilize a real-time database to track histology samples, ensuring accurate and efficient sample management.
Investigators
- Dr. Martin Yaffe, PhD, Principal investigator
- Alison Cheung, PhD, Research Associate
- Gord Mawdsley, BSc., FCCPM, P.Phys., Lab Manager
- James Mainprize, PhD, Research Associate
- Yulia Yerofeyeva, MSc, Project Manager
- Adebayo Adeeko, Technician
- Dan Wang, MSc, Imaging Specialist
- Henry Kim, MLT
BIRL facilities and offices
C748a,b; C750a,b; C727a,b,c; C735a,b – offices
C728 – tissue processing and imaging (x7479)
C729 – tissue microtomy room
C739 – general purpose/meeting room (x5651)
C743 – image digitization/microscopy room
C741 – image device development room
C744 – freezer room/biospecimen storage room
The core services offered by our lab have been developed with support from the Ontario Institute for Cancer Research (OICR) and the Imaging Pipeline Platform.
Research teams in the Imaging Translation Program and the Smarter Imaging Program (formerly known as the One Millimetre Cancer Challenge) are collaborating to improve cancer detection and translate research technologies into the clinic.
Additional links
- News
- More information on CeRIGT sponsors and funding
Key publications
- Pathology Vision (Orlando, annual conference), Nov 2023 – presenting AI method in which immune cell quantifications were linked with the patient survival analysis, an annotated dataset of bladder cancer (Wenchao Han, Dr. Michelle Downes).
- “A modified bleaching method for multiplex immunofluorescence staining of FFPE tissue sections” Dan Wang, Applied Immunohistochemistry & Molecular Morphology, 2024
- A M. Cheung, H Besher, D Wang, K Liu, Y Amemiya, J Chen2, A L Martel, A Seth, M J Yaffe, Integrated image-processing and transcriptomic analysis of cancer-associated fibroblasts (CAFs) in breast cancer subtypes, SPIE, 2023
- Han W, Cheung AM, AL, Ramanathan V, Wang D, Liu K, Yaffe MJ, Martel AM Identification of molecular cell type of breast cancer on digital histopathology images using deep learning and multiplexed fluorescence imaging Medical Imaging 2023: Digital and Computational Pathology. Vol. 12471. SPIE, 2023
- W Han, E Olkhov-Mitsel, A M Cheung, M J Yaffe, M Downes, A L Martel, Cell Quantification Using Machine Learning for Fast Relapse of High Grade Bladder Cancer Patients post-Cystectomy, USCAP, 2023
- W Han, DeepCSeg: a tool kit for instance whole cell and cell nucleus segmentation on immunofluorescence multiplexed images https://github.com/WenchaoHanSRI/DeepCSeg
- Cheung AM, Wang D, Liu K, Hope T, Murray M, Ginty F, Nofech-Mozes S, Martel AL, Yaffe MJ. Quantitative single-cell analysis of immunofluorescence protein multiplex images illustrates biomarker spatial heterogeneity within breast cancer subtypes. Breast Cancer Res. 2021 Dec 18;23(1):114
- Han W, Cheung AM, Yaffe MJ, Martel AL, Cell segmentation for immunofluorescence multiplexed images using two-stage domain adaptation and weakly labeled data for pre-training Scientific Reports volume 12, Art# 4399, 2022
- J Chen, LY Liu, W Han, D Wang, A M Cheung, H Tsui, A L Martel, General stain deconvolution of histopathology images with physics-guided deep learning, bioRxiv, 2022.12. 06.519385
- V Ramanathan, W Han, D Bassiouny, E Rakovitch, A L Martel, Ink removal in whole slide images using hallucinated data, SPIE medical imaging, 2023, Feb, San Diego
- Chen J, Milot L, Cheung HM, Martel AL Unsupervised Clustering of Quantitative Imaging Phenotypes using Autoencoder and Gaussian Mixture Model https://arxiv.org/pdf/1909.02953.pd
- GM Clarke, CMB Holloway, JT Zubovits, S Nofech‐Mozes, K Liu. 2016. Whole‐mount pathology of breast lumpectomy specimens improves detection of tumour margins and focality, Histopathology 69 (1), 35-44
- Clarke GM, Zubovits JT, Shaikh KA, Wang D, Dinn SR, Corwin AD, Santamaria-Pang A, Li Q, Nofech-Mozes S, Liu K, Pang Z, Filkins RJ, Yaffe MJ. Jan2014. A novel, automated technology for multiplex biomarker imaging and application to breast cancer, Histopathology;64(2):242–55
- Sun L, Wang D, Zubovits JT, Yaffe MJ, Clarke GM. 2009. An improved processing method for breast whole-mount serial sections for three-dimensional histopathology imaging. Am J Clin Pathol.; 131:383–92
View all of Dr. Yaffe’s publications on Google Scholar..
Contact
Direct lab number
416-480-6100 ext. 67479