Tissue microarrays: Potential in the Indian subcontinent · advantages of this technology coupled...

Indian Journal of Cancer | January - March 2005 | Volume 42 | Issue 1 5

Tissue microarrays: Potential in the Indiansubcontinent

Venkataraman Girish, Ananthanaranayanan Vijayalakshmi*Department of Pathology, Loyola University Medical Center, Maywood, IL60153 and *Department ofPreventive Medicine, Robert H.Lurie Cancer Center, Northwestern University Feinberg School ofMedicine, Chicago IL60611

Correspondence to: Dr. Girish Venkataraman, E-mail: [email protected]

ReviewArticle

AbstractTissue microarrays (TMAs) are a means of combining hundreds of specimens of tissue on to a single slide for

analysis simultaneously. The evolution of this technology to validate the results of cDNA microarrays has impacted

tremendously in accurately identifying prognostic indicators significant in determining survival demographics for

patients. TMAs can be generated from archival paraffin blocks, combined with sophisticated image analysis software

for reading TMA immunohistochemistry, and a staggering amount of useful information can be generated in terms of

the biomarkers useful in predicting patient outcome. There is a wide range of uses for the TMA technology including

profiling of specific proteins in cancerous tissues and non-cancerous tissues. Given the wide variety of tissue

resources available in India, investment in a dedicated TMA facility will be of immense use in the research arena in

India. This review article discusses the basics of TMA construction, design, the software available for the analysis

of this technology and its relevance to Indian scientists. A potential workflow structure for setting up a TMA facility is

also included.

Key Words: Tissue microarrays, histopathology

Introduction

In the current world of proteomics, the use of highdensity Tissue Microarrays (TMAs)—also called “tissuechips”—has been progressively increasing in the lastdecade ever since its inception in 1998. Arguably, itmight be touted as the most noteworthy developmentin histopathology techniques in the last decade.[1,2]

While DNA microarrays permit expression analysis ofthousands of genes from one tissue specimen on asingle array, TMAs make it possible to analyze hundredsor thousands of tissue specimens in a single experimentusing a single gene or antibody probe. Given thetremendous pace at which novel genes implicated incancer are being discovered using gene chips andexpression microarrays, TMAs hold an immensepotential to validate this genomic data across multipletumor types in a limited time frame. Many pathology

institutions across the world have taken advantage ofthis fact and incorporated TMA facilities exclusively forthis purpose. A thorough understanding of the potentialadvantages of this technology coupled with a unifiedsystem for generating TMA blocks using paraffin blocksfrom high-histopathology-volume institutions in Indiawill help in establishing a good tissue database for theIndian pathologists. The basics of TMA construction,hardware involved and scoring options available arediscussed here.

Basics of TMA construction

TMAs are a method of relocating tissue from standardhistological paraffin blocks such that tissue frommultiple donor blocks can be placed on the samerecipient block. Therefore, the first step in the processof creating a good TMA block is to locate the most

Indian Journal of Cancer | January - March 2005 | Volume 42 | Issue 16

Venkataraman et al: Potential of TMAs

representative area on each H/E slide and carefulidentification of the same area on the correspondingparaffin block of the slide. Once this is done, the blockis positioned underneath the TMA puncher such thatthe representative area is directly under the punchingpins. Thereafter, small “core needle biopsies” of theserepresentative tissues are punched out directly fromdonor paraffin blocks and re-embedded onto a newrecipient TMA paraffin block. Many such cores can beembedded in such a “master-arrayer” block employingthis technique.[2] Using a 0.6 mm diameter puncher,nearly 600 or more tissue cores can be arrayed on astandard glass slide in a precise manner defined by X-Ycoordinates. Although TMA block generation is a labor-intensive and time-consuming process, the ability of thishigh throughput technology to generate data from astaggering number of cases vastly improves statisticalprecision and power.

TMA types, design and data handling options

Having such a large number of spots on one singleslide calls for a precise organization of the tissue spotsat three levels—firstly, in the design and placement ofthe tissue cores in the TMA block; secondly, in thelinking of clinical and pathological data to the correctcore; and lastly, a validated means of scoring andanalyzing any biomarker data so that a meaningfulstatistical analysis can be done.

Types of TMAs

The nature of tissues that can be used on TMAs isvaried, ranging from totally normal tissues from variousorgans to non-neoplastic (like diabetes) and neoplasticones and sometimes even cell lines.[3,4] The CooperativeHuman Tissue Network (CHTN), a division of theNational Cancer Institute (NCI) can provideinvestigators with a wide range of normal tissues placedon a TMA slide so the expression profile of a singleprotein in many tissues can be assessed simultaneously.The information from such TMAs can provide valuableinformation regarding the biology of diseases in whichthese proteins are altered. TMAs based on neoplastictissues also termed as tumor TMAs, are broadlyclassified into three types—multi-tumor arrays,progression arrays (based on stage of tumor) andprognostic arrays where tumors with known clinical endpoints are arrayed.[5-7]

Design of TMAs

An integral part of constructing TMAs is the design.After deciding on the optimum number of cases with

sufficient clinical data, an important issue that meritsconsideration is the number of cores per case to putinto the TMA block while evaluating any biomarkerexpression by immunohistochemistry. Although thisnumber is variable, for most tumor TMAs, three tofour cores/block adequately represent a biomarker’sability to predict survival outcomes.[8] Often two coresare taken from the donor block; one core is taken fromthe center of the donor block and the other close tothe periphery of the block. It is important to includeappropriate positive and negative controls besidesorientation cores. It would be useful to chart out anMS Excel worksheet identifying the numbers of thevarious cores in the same pattern in which the cores areembedded in the TMA block. In addition, inclusion ofirrelevant tissue at defined X-Y coordinates can help inaccurate orientation of the cores. Often cores fromtumors of same T size, stage or grade or tissue oforigin can be clustered together on the TMA block. Atypical low-density TMA block and slide is shown inFigure 1. Last but not the least, it is important toensure that all blocks used for TMA construction havebeen fixed and processed similarly because someimmunohistochemical markers may not work when thetissue is fixed in a different fixative.

The technology used to generate TMA can range frommanual to fully automated systems. One of the moreprominent companies include Beecher Instruments (SanPrairie, Wisconsin, USA), which features manual, semi-automated as well as fully automated systems. The costcan vary between $10,000 -$42,000 (USD) dependingon the type of the equipment. Cost-effective alternativemethods of generating TMA blocks can also be adapted

Figure 1: A typical low-density TMA block containing 30 cores(0.6mm each). The corresponding H/E slide with appropriate ID isdepicted. Investigators musts be aware that individual cores canget depleted and the hence the need for duplicate representationof each core on the TMA block. Orientation cores have not beenincluded here because of the low-density of cores on the block



if 60 or fewer number of cores are to be imprinted intothe TMA block. A standard microscope fitted with aholing needle or a blunted 16G bone marrow trephineneedle can be adapted to design TMA blocks at afraction of the price of commercial instruments.[9]

Data acquisition, analysis and integration

Digital scanning and analysis

The next issue of importance is reading the TMA spotsin an orderly, reproducible and reliable manner. Ofcourse, the intuitive option would be to read one coreafter another manually under a bright-field microscope.Even so, keeping track of the precise position of eachspot can become a painstaking process if there are morethan 300 tissue spots on any single slide. Consequently,digital options are popular for analyzing biomarkerexpression on TMA. Systems like the BLISS Imaging(Bacus Labs, Lombard, IL) and ACIS system ofChromavision (www.chromavision.com) can scan andacquire images of all spots on the slide in one gowithout the user having to manually focus each spotseparately before acquiring. Tracking and recording thescores from these tissue cores can be done usingcommercially available proprietary software. BacusLaboratories also markets a TMA analysis softwarecalled ‘TMAscore’. With this, TMA slides are scannedinto virtual slides and placed on a network. Thereafter,multiple people can collaborate, and score each coremanually or by using the aforementioned proprietarysoftware.

Cost-effective alternatives for analysis

Given the logistics involved in using proprietarysoftware, there is a need for exploring other economicalmeans of doing spot scoring and analysis. Somenoteworthy cost-effective software alternatives areavailable for recording patient data, histology details andcoordinates of each core for scoring, some of which aredescribed below. This still requires an image-capturingsystem to create an archived database of images forfurther analysis. The analysis can however be doneusing an ‘open-source’ free software available over theInternet for academic use.

In the first of these, a novel relational database forTMA analysis has recently been described using AdobePhotoshop (an imaging software) for image-editing andMicrosoft Excel for recording core coordinates andperforming scoring.[10] The authors of this study explainthat a library of digital images of individual cores isstored in a specified folder of the computer. An Excel

worksheet is created and grid locations of the cores arenoted in rows and columns in the same pattern as inthe TMA slide. Thereafter, hyperlinks are inserted foreach grid such that clicking on the hyperlinkcorresponding to a specific core in the Excel worksheetwill open the corresponding image of the same core forscoring. This method is especially useful if the numberof cores on a slide is not going to exceed 150 cores.

For managing high-density TMAs with more than 600cores per slide, an excellent set of software tools forhigh-throughput analysis has been developed at StanfordUniversity.[11] These investigators used the BLISSSystem mentioned previously for generating a databaseof images. Immunohistochemical staining results arerecorded into an MS Excel Worksheet and this Exceldata is re-formatted by a program called “TMA-Deconvoluter” which converts the Excel data into a textfile so that a Hierarchical Cluster Analysis can be doneusing the “Cluster” and “TreeView” software—thisanalyzes the relatedness within tumor subsets dependingon the immunohistochemical biomarker profile. TheTreeView software generates dendrograms akin to theones seen with cDNA microarray data. Free access toTMA-Deconvoluter is possible at the Stanford TMAWebsite (http://genome-www.stanford.edu/TMA) whilethe other two software programs are available at thewebsite of Michael B. Eisen’s lab (http://rana.lbl.gov/EisenSoftware.htm).

Another noteworthy open-source java-based softwarecalled “TMAJ” is available from the website of theJohns Hopkins University TMA core facility (http://tmaj.pathology.jhmi.edu/). This helps in recordingpathology data as well core tracking and scoring. Alicense is however required for users with potentialcommercial interests.

Data integration

As long as TMA slides are read by a single pathologist,it is possible to manage data with just MS Excel andstatistical analysis software. When multiple investigatorsread the same slides, systematic data integration canbecome overwhelming. TMAs generate an immenseamount of data and so collaborating pathologistsanalyzing the same TMAs in different institutions needto be aware that they have to record the generated datain a uniform and standardized format. This is crucialbecause Excel data files generated by multiplepathologists across the country scoring the same TMAslides, can be integrated for statistical analysis only ifthe scoring format is uniform between all theinvestigators. Addressing this issue, a TMA data


exchange software specification language for free use hasbeen developed in XML (eXtensible Markup Language)recently.[12,13] The basic pre-requisite of this dataintegration language is that all investigators need tohave their Excel data scoring files with the samecolumn headings viz. Patient ID, Age, core ID etc inthe same order. A PERL (Practical Extraction andReport Language) (another computer language) scriptwill export all these Excel data in the final XMLdatabase accurately only if the column headings are thesame in all Excel files. If XML is difficult, collaboratingpathologists can also use MS Excel exclusively andconfigure it to concatenate data from multipleinvestigators for analysis.[14]

Advantages, validation and caveats of TMAs

There are a number of plus points to using TMAsinstead of standard histology sections. Firstly, a largenumber of cases can be analyzed simultaneously andthis in turn, improves the power and precision of anystatistical analysis considerably. Schraml et al analyzedthe oncogenes CCD1, CMYC and ERBB2 in 397tumor spots by Fluorescent In-Situ Hybridization injust one week attesting to the rapidity with whichTMAs can generate results.[15]

Secondly, immunohistochemistry on a single TMA slideis much more time-saving and economical in terms ofthe amount of antibodies and reagents used ascompared to paraffin sections while ensuring uniformtreatment of all cores on a slide—no batch to batchvariability needs to be accounted for as with standardsections.[16] Van de Rijn et al have described the utilityof a 29-case mini TMA of breast cancer blocks as auseful control for Estrogen Receptor. Instead of a singlestrong positive control slide, the TMA includes aspectrum of positivity from weak to strong.[17] As aresult, any change in the strength of the detection canbe easily detected. Besides, inter-laboratory qualitycontrol for immunohistochemistry can also be assessed.

On the other hand, TMAs can be used to test newprognostic candidate biomarkers evolving from cDNAmicroarray studies. Natkunam et al were able to test anew marker of plasmacytic differentiation calledMUM1/RF4 against 1335 different human tumors anddemonstrate it to be a sensitive albeit non-specificmarker to the extent that even melanocytic tumorsstained positive.[18]

Yet another study has demonstrated the elegant way inwhich TMAs validated the results of a cDNAmicroarray analysis.[19] The authors screened 5184 genes

and discovered 2 genes namely, IGFBP2 and HSP27that were upregulated in xenografts of hormonerefractory prostate cancer (HRPC). They followed thisup with immunohistochemistry for both markers on aTMA and found that 100% of HRPC cases werepositive for IGFBP2 while none of the normal prostatecores were positive for the same marker.

Criticism is often directed towards the fact that thesmall size of TMA cores may not adequately representbiomarkers (like ER) that are distributedheterogeneously in the tissue. Addressing this problemof spatial heterogeneity, Camp et al found that just two0.6 mm cores from the same block provided equivalentinformation on ER indices as that of whole sections.[20]

Thus, intra-tumoral heterogeneity will not be an issueof significant concern.

Caveats with TMAs: With TMAs, investigators alsoneed to be able to decide if constructing a TMA inany situation is going to give additional informationfor the extra price being paid; if you have only 30donor blocks of a prostatic adenocarcinoma, it mightwork out better doing conventional one slide-onesection IHC rather than constructing a TMA. Thecost-benefit ratio is high only if the array is valuable(i.e. if it contains rare to find tissues and lesions),the number of cores on the array is very high or ifthere is a need to assess many proteins at one time.If too many spatially close cores (>6) are marked forpunching on the same block, the technologist willhave a hard time locating these on the block. Lastly,it is necessary to do a Quality Assurance (QA) withevery TMA slide to see if the representative tissue ispresent at the pre-defined X-Y coordinate on the slide.The representative tissue of interest may often getdepleted in deeper sections (e.g. breast lobules orareas of HGPIN in blocks from radicalprostatectomy). In such cases, an auxiliary TMAblock using the same tissues should be constructedusing larger tissue cores.

Resources: The National Cancer Institute (NCI,Bethesda, MD USA) has set up a Tissue ArrayResearch Program, TARP (http://ccr.cancer.gov/tech_initiatives/tarp/) to facilitate access to multi-tumorTMA slides to investigators interested in procuring pre-designed TMA slides. There are two other divisions ofthe NCI namely, CBCTR and CPCTR (Central Breast/Prostate Cancer Tissue Resource databases respectively)which have a repository of pre-arrayed TMA slides.Access to TMA slides containing prostate and breastcancer cores is available from them at a nominal priceafter approval of the proposed intent of use.



On the other hand, commercial companies will alsomake TMA blocks if they are provided with thestandard paraffin blocks (www.petagen.com) althoughthe costs can get prohibitive if high-density arrays areplanned. In a recent article, Mobasheri et al haveprovided a comprehensive list of the commercialcompanies constructing TMAs.[21]

Potential of TMA in India

Given the immense potential of this technology, it isclearly evident that TMAs will play a major role in theforeseeable future in the arena of evaluating potentialbiomarkers. There are a number of high-volumehistopathology institutions within India like AIIMS,PGIMER, JIPMER and Tata Memorial Hospital. Anyone of these institutes can operate a ‘TMA core facility’.All other institutes could send in their blocks alongwith relevant clinico-pathological information for TMAconstruction. In return, these institutes would then haveaccess to the TMA slides generated from the blocks at anominal price that would hopefully cover for the pay ofthe technicians and computer programmers working atthe core facility.

Two technical staff (one for receiving and the other forarraying blocks), one computer programmer (forcomputerized cataloguing of the TMA database andgenerating template MS Excel data files) and onepathologist for marking out areas to be punched aresufficient within such a core facility. Figure 2 depicts apotential hierarchy to streamline the process in ourcountry. Sample forms that will be useful for recordingarray design (Form-A), clinico-pathological data of

blocks received from the peripheral centers (Form-B);Form-C and D for scoring recording the clinico-pathologic data of the array. All outgoing forms can beincluded as multiple Excel worksheets in one singleExcel file that makes data handling and integration aneasier job. A useful thing for pathologists participatingin such TMA facilities would be to learn to use MSExcel in a proficient manner so that they can recorddata efficiently. For a start, programmers can design theaforementioned Excel guide sheets based on theCHTNs Excel sheets available from their website http:// f a c u l t y. v i r g i n i a . e d u / c h t n - t m a / 2 0 0 2 N 1 /CHTN2002N1X.xls.

On the logistic front, a manual tissue microarrayer oftencomes at the price of a high-end microtome and themonetary savings from using TMAs will definitely offsetthe initial costs of establishment, in the long run. Forpartial funding of core center technologists, 20% of theTMA slides from tumor arrays may be allocated toforeign researchers at a nominal price to generaterevenue.

TMA slides with relevant information could be used asthesis material by PhD and post-graduate pathologyresidents who are required to complete a thesis work inpartial fulfillment of their residency program. Currently,pathology residents often have a difficult time accessingarchival paraffin blocks, slides and clinical informationin the process of working on their thesis. Unfortunately,many medical and even pathology postgraduates areoften ignorant about this technology. Students, facultyand technicians at all levels should be educated aboutTMA technology. At the medical school level, thepathology curriculum should probably try to include achapter on recent medical technology like TMA andothers that have impacted significantly on the researchfront. At the postgraduate level, invited talks by expertsin the field at national conferences like the IAPM wouldbe a good way to introduce this technology at all levelsin the scientific community. Another potential advantageis that a single TMA slide with all possible grades ofany single tumor type (like prostate cancer) will helppathology post-graduates in mastering the fine art oftumor grading. This in turn will help in reducing inter-observer variability while grading these tumors in dailysurgical pathology practice.

On the other hand, there will be abundant material forthe faculty to conduct research without having to useup large amounts of costly antibodies. On the researchfront, there will be ample opportunities to collaboratewith overseas institutions where access to histologytissues is relatively difficult owing to bureaucratic issues.

Figure 2: Flowchart depicting the flow of blocks and slides ina potential TMA setup in order, going through steps 1-5. ThePeripheral centres send in their blocks with relevant clinicalinformation to the Core Centre. The Core Centre designs thearray and creates TMA blocks and slides. It is important for acoordinating computer programmer at the Core Centre todesign standardized forms for receipt of blocks, clinical dataand array design

PeripheralCentre

Send Blocks AND Clinical Data

On request

Archive blocks and unstained slides

1

2

34

5

Unstained SlidesAttach Relevant Excel

Forms

FilledExcel Forms-C fromPeripheral Centres

CORE CENTRE

PATHOLOGIST TECHS (x2) PROGRAMMER

• Create ExcelForms A-D

• Integrate XMLdata for statisticalanalysis

• Block Receipt• ID Assigning/Label• TMA Block/slidegenerationfrom Both inputs

• Array Design

• Mark Areas forcoring

• Analyze integrateddata withBiostatistician



The ample autopsy material available within India canserve as a valuable source of tissues for TMA blockswhile collaborating foreign institutions can supplyprimary antibodies thus pooling resources.

Final word

TMA technology is a powerful proteomic platform thatplays an integral role in understanding proteinexpression patterns in a wide variety of normal andneoplastic tissues. The technology is highly evolved andhas abundant potential to bridge the chasm betweentranslational research and clinical therapeutics. Given thewealth of tissue resources available in our country, itonly seems that the adopting this state-of-the-arttechnology will go a long way in benefiting studentsand researchers alike in the time to come.

Addendum: All the hyperlinks have been tested and thewebsites are accessible as of December 29, 2004.

References

1. Wan WH, Fortuna MB, Furmanski P. A rapid and efficient methodfor testing immunohistochemical reactivity of monoclonal anti-bodies against multiple tissue samples simultaneously. J ImmunolMethods 1987;103:121-9.

2. Kononen J, Bubendorf L, Kallioniemi A, Barlund M, Schraml P,Leighton S, et al. Tissue microarrays for high-throughput molecu-lar profiling of tumor specimens. Nat Med 1998;4:844-7.

3. Mobasheri A, Marples D. Expression of the AQP-1 water chan-nel in normal human tissues: A semiquantitative study using tis-sue microarray technology. Am J Physiol Cell Physiol2004;286:C529-37.

4. Moskaluk CA, Stoler MH. Agarose mold embedding of culturedcells for tissue microarrays. Diagn Mol Pathol 2002;11:234-8.

5. Nocito A, Kononen J, Kallioniemi OP, Sauter G. Tissue microarrays(TMAs) for high-throughput molecular pathology research. Int JCancer 2001;94:1-5.

6. Shergill IS, Arya M. Tissue microarrays. Expert Rev Mol Diagn2004;4:421-3.

7. Shergill IS, Shergill NK, Arya M, Patel HR. Tissue microarrays:A current medical research tool. Curr Med Res Opin

2004;20:707-12.8. Rubin MA, Dunn R, Strawderman M, Pienta KJ. Tissue microarray

sampling strategy for prostate cancer biomarker analysis. Am JSurg Pathol 2002;26:312-9.

9. Hidalgo A, Pina P, Guerrero G, Lazos M, Salcedo M. A simplemethod for the construction of small format tissue arrays. J ClinPathol 2003;56:144-6.

10. Shaknovich R, Celestine A, Yang L, Cattoretti G. Novel relationaldatabase for tissue microarray analysis. Arch Pathol Lab Med2003;127:492-4.

11. Liu CL, Prapong W, Natkunam Y, Alizadeh A, Montgomery K, GilksCB, et al. Software tools for high-throughput analysis and archiv-ing of immunohistochemistry staining data obtained with tissuemicroarrays. Am J Pathol 2002;161:1557-65.

12. Berman JJ, Edgerton ME, Friedman BA. The tissue microarray dataexchange specification: A community-based, open source toolfor sharing tissue microarray data. BMC Med Inform Decis Mak2003;3:5.

13. Berman JJ, Datta M, Kajdacsy-Balla A, Melamed J, Orenstein J,Dobbin K, et al. The tissue microarray data exchange specifica-tion: Implementation by the Cooperative Prostate Cancer TissueResource. BMC Bioinformatics 2004;5:19.

14. Manley S, Mucci NR, De Marzo AM, Rubin MA. Relational data-base structure to manage high-density tissue microarray data andimages for pathology studies focusing on clinical outcome: Theprostate specialized program of research excellence model. AmJ Pathol 2001;159:837-43.

15. Schraml P, Kononen J, Bubendorf L, Moch H, Bissig H, Nocito A,et al. Tissue microarrays for gene amplification surveys in manydifferent tumor types. Clin Cancer Res 1999;5:1966-75.

16. Milanes-Yearsley M, Hammond ME, Pajak TF, Cooper JS, ChangC, Griffin T, et al. Tissue micro-array: A cost and time-effectivemethod for correlative studies by regional and national cancerstudy groups. Mod Pathol 2002;15:1366-73.

17. van de Rijn M, Gilks CB. Applications of microarrays to histopa-thology. Histopathology 2004;44:97-108.

18. Natkunam Y, Warnke RA, Montgomery K, Falini B, van De Rijn M.Analysis of MUM1/IRF4 protein expression using tissuemicroarrays and immunohistochemistry. Mod Pathol2001;14:686-94.

19. Bubendorf L, Kolmer M, Kononen J, Koivisto P, Mousses S, ChenY, et al. Hormone therapy failure in human prostate cancer: Analy-sis by complementary DNA and tissue microarrays. J Natl CancerInst 1999;91:1758-64.

20. Camp RL, Charette LA, Rimm DL. Validation of tissue microarraytechnology in breast carcinoma. Lab Invest 2000;80:1943-9.

21. Mobasheri A, Airley R, Foster CS, Schulze-Tanzil G, Shakibaei M.Post-genomic applications of tissue microarrays: Basic research,prognostic oncology, clinical genomics and drug discovery. HistolHistopathol 2004;19:325-35.


Tissue microarrays: Potential in the Indian subcontinent · advantages of this technology coupled...

Documents

Transcript of Tissue microarrays: Potential in the Indian subcontinent · advantages of this technology coupled...