PRACTICE EVOLUTION: Decentralized Computer-Assisted IHC Image Analysis Liron Pantanowitz, MD, FCAP...
-
Upload
gregory-daniels -
Category
Documents
-
view
217 -
download
0
Transcript of PRACTICE EVOLUTION: Decentralized Computer-Assisted IHC Image Analysis Liron Pantanowitz, MD, FCAP...
PRACTICE EVOLUTION: Decentralized Computer-Assisted IHC Image Analysis
Liron Pantanowitz, MD, FCAPDirector of Pathology Informatics
Richard C. Friedberg, MD PhD, FCAPChairman, Department of Pathology
Baystate Health, Springfield, MATufts University School of Medicine
Why Are We Doing This?Why Are We Doing This?
• Practice BackgroundPractice Background• Today’s EnvironmentToday’s Environment
Increased technological innovationIncreased technological innovation Increased biological informationIncreased biological information Increased clinical demandIncreased clinical demand
• Convergence of two independent long Convergence of two independent long term trendsterm trends
Key Trend #1 in the Practice of Key Trend #1 in the Practice of Anatomic PathologyAnatomic Pathology
• Evolution along Clinical Pathology linesEvolution along Clinical Pathology lines Greater concern with analytical precision, Greater concern with analytical precision,
reproducibility, accuracy, specificity, reliabilityreproducibility, accuracy, specificity, reliability Qualitative becoming quantitativeQualitative becoming quantitative ““Stains” becoming “assays”Stains” becoming “assays” Results directly tied to treatment, not just prognosisResults directly tied to treatment, not just prognosis Diminishing “guild” mentality with anointed expertsDiminishing “guild” mentality with anointed experts
• ExamplesExamples IHC & ELISAIHC & ELISA Her2/neu & Herceptin Her2/neu & Herceptin
Key Trend #2 in the Practice of Key Trend #2 in the Practice of Anatomic PathologyAnatomic Pathology
• Evolution along Radiology/Imaging linesEvolution along Radiology/Imaging lines Analog images establish the fieldAnalog images establish the field Market & technology forces start trend to digital Market & technology forces start trend to digital
imaging imaging • Initially, scanning of analog imagesInitially, scanning of analog images• Later, digitally acquired imagesLater, digitally acquired images
Digitalization of images allows new applicationsDigitalization of images allows new applications Significant workload & throughput implications Significant workload & throughput implications
• ExamplesExamples PACSPACS Convergence imagingConvergence imaging WindowingWindowing Dynamic imagesDynamic images Telediagnostics Telediagnostics
ExpectationsExpectations
• EventuallyEventually Every “image-based” pathologist will use Every “image-based” pathologist will use
computer-assisted analytic tools to assay computer-assisted analytic tools to assay specimensspecimens
Intelligently designed PACS will revolutionize Intelligently designed PACS will revolutionize pathology workflowpathology workflow
Increased reliance upon pathology Increased reliance upon pathology
Breast Cancer & Breast Cancer & Immunohistochemistry (IHC)Immunohistochemistry (IHC)
Determining breast tumor markers (ER, PR & HER-2/neu) for Determining breast tumor markers (ER, PR & HER-2/neu) for prognostic & predictive purposes by IHC &/or FISH is the prognostic & predictive purposes by IHC &/or FISH is the standard of practice.standard of practice.
IHC score/quantification by manual microscopy is currently IHC score/quantification by manual microscopy is currently accepted as the traditional gold standard.accepted as the traditional gold standard.
Surgical Pathology workflow involves:Surgical Pathology workflow involves:• Pre-analytic preparationPre-analytic preparation (e.g. tissue fixation & processing) (e.g. tissue fixation & processing)• AnalysisAnalysis (i.e. staining of controls & patient slides) (i.e. staining of controls & patient slides)• Post-analytical componentPost-analytical component (e.g. quantification & reporting) (e.g. quantification & reporting)
DiscrepanciesDiscrepancies between HER2 IHC & FISH mainly reflect errors between HER2 IHC & FISH mainly reflect errors in manual interpretation & not reagent limitations (in manual interpretation & not reagent limitations (Bloom & Bloom & Harrington. AJCP 2004; 121:620-30Harrington. AJCP 2004; 121:620-30).).
Inter- & intra-observer differences in scoringInter- & intra-observer differences in scoring occur: occur:• Most notably with borderline & weakly stained casesMost notably with borderline & weakly stained cases• Related to fatigue & subjectivity of human observersRelated to fatigue & subjectivity of human observers
Accuracy is RequiredAccuracy is Required Accuracy = the amount by which a measured Accuracy = the amount by which a measured
value adheres to a standard. value adheres to a standard. The need for precise ER, PR & HER2/neu status in The need for precise ER, PR & HER2/neu status in
breast cancer is required to ensure appropriate breast cancer is required to ensure appropriate therapeutic intervention.therapeutic intervention.
Lay press have communicated concerns over Lay press have communicated concerns over inaccuracies in breast biomarker testing.inaccuracies in breast biomarker testing.
Threat of having to refer such testing to reference Threat of having to refer such testing to reference laboratories.laboratories.
Is computer assisted image analysis (CAIA) a Is computer assisted image analysis (CAIA) a better (i.e. more accurate & reproducible) method better (i.e. more accurate & reproducible) method for scoring IHC?for scoring IHC?
GuidelinesGuidelines ASCO/CAP Guideline Recommendations for HER2/neu testing in ASCO/CAP Guideline Recommendations for HER2/neu testing in
breast cancerbreast cancer ( (Wolff et al. Arch Pathol Lab Med 2007; 131:18Wolff et al. Arch Pathol Lab Med 2007; 131:18))• Image analysis can be an effective tool for achieving consistent Image analysis can be an effective tool for achieving consistent
interpretationinterpretation• A pathologist must confirm the image analysis resultA pathologist must confirm the image analysis result• Image analysis equipment (including optical microscopes) must be Image analysis equipment (including optical microscopes) must be
calibrated, subjected to regular maintenance & internal QC evaluationcalibrated, subjected to regular maintenance & internal QC evaluation• Image analysis procedures must be validatedImage analysis procedures must be validated
Canadian National Consensus Meeting on HER2/neu testing in Canadian National Consensus Meeting on HER2/neu testing in breast cancerbreast cancer ( (Hanna et al. Current Oncology 2007; 14:149-53Hanna et al. Current Oncology 2007; 14:149-53))• Use of image analysis systems can be useful to enhance reproducibility Use of image analysis systems can be useful to enhance reproducibility
of scoringof scoring• Pathologists must supervise all image analysesPathologists must supervise all image analyses
FDA clearance FDA clearance for CAIA in vitro diagnostic use of HER-2/neu, ER, and PR for CAIA in vitro diagnostic use of HER-2/neu, ER, and PR IHC has been obtained by several companiesIHC has been obtained by several companies
CAIA vs. Manual ScoreCAIA vs. Manual ScoreRemmele & Schicketanz. Pathol Res Pract 1993; 189:862-6Remmele & Schicketanz. Pathol Res Pract 1993; 189:862-6
““Subjective grading of slides is a simple, Subjective grading of slides is a simple, rapid and useful method for the rapid and useful method for the determination of tissue receptor content determination of tissue receptor content and must not be replaced by expensive and must not be replaced by expensive and time-consuming computer-assisted and time-consuming computer-assisted image analysis in daily practice.”image analysis in daily practice.”
Data on CAIA & IHCData on CAIA & IHC Early studies showed CAIA was no better than visual analysisEarly studies showed CAIA was no better than visual analysis
((Schultz et al. Anal Quant Cytol Histol 1992; 14:35-40Schultz et al. Anal Quant Cytol Histol 1992; 14:35-40)) Few studies have shown that manual & CAIA are comparable Few studies have shown that manual & CAIA are comparable
(Diaz et al. Ann Diagn Pathol 2004; 8:23-7)(Diaz et al. Ann Diagn Pathol 2004; 8:23-7) Most studies found CAIA to be superior to manual methods Most studies found CAIA to be superior to manual methods
(Taylor & Levenson. Histopathology 2006; 49::411-24; McClelland et al. (Taylor & Levenson. Histopathology 2006; 49::411-24; McClelland et al. Cancer Res 1990; 50:3545-50; Kohlberger et al. Anticancer Res 1999; Cancer Res 1990; 50:3545-50; Kohlberger et al. Anticancer Res 1999; 19:2189-93; Wang et al. Am J Clin Pathol 2001; 116:495-503; Turner et al. 19:2189-93; Wang et al. Am J Clin Pathol 2001; 116:495-503; Turner et al. USCAP 2008 abstract 1694).USCAP 2008 abstract 1694).• Provides effective qualitative & quantitative evaluationProvides effective qualitative & quantitative evaluation• More consistent than manual & digital microscopyMore consistent than manual & digital microscopy• More precise (scan per scan) than pathologistsMore precise (scan per scan) than pathologists
One study showed agreement between different CAIA One study showed agreement between different CAIA systems:systems: Chroma Vision ACIS & Applied Imaging Ariol SL-50 Chroma Vision ACIS & Applied Imaging Ariol SL-50 (Gokhale et al. Appl Immunohistochem Mol Morphol 2007; 15:451-5)(Gokhale et al. Appl Immunohistochem Mol Morphol 2007; 15:451-5)
Published ConsiderationsPublished Considerations Expense of CAIA may be hard to justify where Expense of CAIA may be hard to justify where
volumes are lowvolumes are low Image analysis frequently requires interactive Image analysis frequently requires interactive
input by the pathologistinput by the pathologist Increased time requirements Increased time requirements Systems may be discrepant when tumor cells Systems may be discrepant when tumor cells
have low levels of staininghave low levels of staining Interfering non-specific staining within selected Interfering non-specific staining within selected
areasareas Images must be free from artifactsImages must be free from artifacts Small amounts of stained tissue can erroneously Small amounts of stained tissue can erroneously
generate lower scoresgenerate lower scores
CAIA SystemsCAIA Systems
• ImageJImageJ (NIH developed freeware) (NIH developed freeware)• Adobe PhotoshopAdobe Photoshop software software
(Lehr et al J Histochem Cytochem 1997; 45:1559-65)(Lehr et al J Histochem Cytochem 1997; 45:1559-65)
• Automated Cellular Imaging SystemAutomated Cellular Imaging System (Chroma Vision)(Chroma Vision)
• Pathiam Pathiam (BioImagene)(BioImagene)• Applied Imaging AriolApplied Imaging Ariol (Gentix Systems) (Gentix Systems)• SpectrumSpectrum (Aperio) (Aperio)
Image Analysis & AlgorithmsImage Analysis & Algorithms Object-Oriented Image AnalysisObject-Oriented Image Analysis (morphology- (morphology-
based)based) Involves color normalization, background Involves color normalization, background
extraction, segmentation, classification & feature extraction, segmentation, classification & feature selectionselection
Separation of tissue elementsSeparation of tissue elements (e.g. tumor (e.g. tumor epithelium) from background (e.g. stroma) epithelium) from background (e.g. stroma) permits selection of areas of interest & filtering permits selection of areas of interest & filtering out of unwanted areasout of unwanted areas
Region of Interest (ROI)Region of Interest (ROI) is subject to further is subject to further image analysis (computation of diagnostic score)image analysis (computation of diagnostic score)
QuantificationQuantification of results of results
Digital AlgorithmDigital Algorithm
Courtesy of BioImagene
Courtesy of BioImagene
Courtesy of BioImagene
Validation & Implementation Validation & Implementation at Baystate Healthat Baystate Health
Distant medical centersDistant medical centers Significant breast IHC caseloadSignificant breast IHC caseload Need to mimic daily practiceNeed to mimic daily practice
• avoid central (single user) image analysisavoid central (single user) image analysis Bandwidth limitationsBandwidth limitations Whole slide imager availabilityWhole slide imager availability Professional reluctance to read digital Professional reluctance to read digital
imagesimages
Key ComponentsKey Components
Multimedia PC upgradeMultimedia PC upgrade Spot Diagnostic digital cameras for each Spot Diagnostic digital cameras for each
workstation workstation PathiamPathiam (BioImagene) web-based (BioImagene) web-based
applicationapplication Server (Oracle database + application + Server (Oracle database + application +
image file storage)image file storage) Training & ValidationTraining & Validation
WORKFLOWWORKFLOW
CONTROL IHC
PATIENT IHC
FOV ANALYSIS
REPORT GENERATION
NEED FOR STANDARDIZATIONNEED FOR STANDARDIZATION
Calibrated WorkstationsCalibrated Workstations
FOV IHC AnalysisFOV IHC Analysis FFPE breast cases routinely stained for ER, PR & HER2-neuFFPE breast cases routinely stained for ER, PR & HER2-neu Standardized camera acquisition settings (calibration)Standardized camera acquisition settings (calibration) Pathologists (n=3) acquired 3-5 FOVs (each at 20x Mag.)Pathologists (n=3) acquired 3-5 FOVs (each at 20x Mag.) Uniform jpg image file formats used (4 Mb)Uniform jpg image file formats used (4 Mb) Post-processing image manipulation was avoidedPost-processing image manipulation was avoided Control parameter set defined/IHC run (default/modified)Control parameter set defined/IHC run (default/modified) ER/PR nuclear staining analyzed using the Allred scoring ER/PR nuclear staining analyzed using the Allred scoring
system (i.e. proportion + intensity score = TS)system (i.e. proportion + intensity score = TS) HER-2/neu membranous staining evaluated per ASCO/CAP HER-2/neu membranous staining evaluated per ASCO/CAP
2007 recommendations (0, 1+, 2+, 3+)2007 recommendations (0, 1+, 2+, 3+) Manual vs. CAIA comparison tracked (IHC score, time & Manual vs. CAIA comparison tracked (IHC score, time &
problems)problems) FISH for HER2/neu obtained on several casesFISH for HER2/neu obtained on several cases
ER/PR Correlation (N=29)ER/PR Correlation (N=29)
Bio- markerBio- marker ConcordantConcordant
CasesCasesDiscordant Discordant
CasesCases
ER+ER+ 1616 00
ER -ER - 44 2*2*
PR+PR+ 1414 00
PR -PR - 44 3*3*
* 3 cases
HER-2/Neu Results (N=28)HER-2/Neu Results (N=28)
ScoreScore 0/1+0/1+ 2+2+ 3+3+
0/1+0/1+ 1616 1*1*
2+2+ 33 1**1**
3+3+ 44
CA
IA Manual Scoring
FISH RESULTS:
* Negative (Ratio 1.04)
** Abnormal (Ratio 6.5)
HER-2/Neu FISH CorrelationHER-2/Neu FISH CorrelationManual ScoreManual Score CAIA ScoreCAIA Score FISH ResultFISH Result
00 00 Negative (1.06)Negative (1.06)
00 00 Negative (0.93)Negative (0.93)
11 11 Negative (1.04)Negative (1.04)
11 00 Negative (1.00)Negative (1.00)
11 00 Negative (1.07)Negative (1.07)
11 00 Negative (1.66)Negative (1.66)
22 11 Negative (1.04)Negative (1.04)
33 22 Abnormal (6.5)Abnormal (6.5)
Challenging CasesChallenging Cases
Infiltrating Lobular Carcinoma Cytoplasmic StainingInfiltrating Lobular Carcinoma Cytoplasmic Staining
Lessons LearnedLessons Learned
Decentralized CAIA for IHC designed to Decentralized CAIA for IHC designed to mimic daily surgical pathology workflow in mimic daily surgical pathology workflow in practice is feasiblepractice is feasible
Image acquisition requires standardizationImage acquisition requires standardization Tissue heterogeneity may impact FOV Tissue heterogeneity may impact FOV
selection (whether biological or due to IHC selection (whether biological or due to IHC variation)variation)
Pathologists must supervise CAIA systemsPathologists must supervise CAIA systems
Future ProspectsFuture Prospects Adopt virtual workflow-centric systems feasible for routine Adopt virtual workflow-centric systems feasible for routine
practice (that may potentially show better results)practice (that may potentially show better results)• E.g. Whole slide imaging (WSI) to eliminate the need to E.g. Whole slide imaging (WSI) to eliminate the need to
standardize different systemsstandardize different systems Automatic ROI selection & image analysisAutomatic ROI selection & image analysis Shortened analysis timeShortened analysis time AP-LIS & CAIA system integrationAP-LIS & CAIA system integration
• To improve workflowTo improve workflow• Permit disparate systems to access the same digital images & Permit disparate systems to access the same digital images &
case datacase data Learning algorithmsLearning algorithms
• Systems that improve with experience following pathologist Systems that improve with experience following pathologist feedbackfeedback
Clinical outcome studies are neededClinical outcome studies are needed• In one study, CAIA for ER IHC yielded results that did not differ from In one study, CAIA for ER IHC yielded results that did not differ from
human scoring against patient outcome gold standards (human scoring against patient outcome gold standards (Turbin et al. Turbin et al. Breast Cancer Res Treat 2007Breast Cancer Res Treat 2007))
AcknowledgementsAcknowledgements
Christopher N. Otis, MDChristopher N. Otis, MD
Giovana M. Crisi, MDGiovana M. Crisi, MD
Andrew Ellithorpe, MHSAndrew Ellithorpe, MHS
Peter Marquis, BAPeter Marquis, BA
BioImageneBioImagene
TRANSFORMING PATHOLOGY:Emerging technology driving practice
innovation