Quantitative evaluation of protein expression as a function of tissue microarray core diameter: is a large (1.5 mm) core better than a small (0.6 mm) core?
* Context.--Tissue microarrays (TMAs) have emerged as a
high-throughput technology for protein evaluation in large cohorts. This
technique allows maximization of tissue resources by analysis of
sections from 0.6-mm to 1.5-mm core "biopsies" of standard
formalin-fixed, paraffin-embedded tissue blocks and by the processing of
hundreds of cases arrayed on a single recipient block in an identical
Objective.--To assess the expression of a series of biomarkers as a function of core size. Although pathologists frequently feel better if larger core sizes are used, there is no evidence in the literature showing that large cores are better (or worse) than small cores for assessment of TMAs.
Design.--Estrogen receptor, HER2/neu, epidermal growth factor receptor, STAT3, mTOR, and phospho-p70 S6 kinase were measured by immunofluorescence with automated quantitative analysis. One random 0.6-mm field (one 0.6-mm spot) was compared to 6 to 12 fields per spot, representing 1-mm and 1.5-mm cores, for 3 different tumor types.
Results.--We show that measurement of a single random 0.6-mm spot was comparable to analysis of the whole 1mm or 1.5-mm spot (Pearson R coefficient varying from 0.87-0.98) for all markers tested.
Conclusions.--Since TMA technology is now being used in all phases of biomarker development, this work shows that TMAs with 0.6-mm cores are as representative as those with any common larger core size for optimization of standardized experimental conditions. Given that a greater number of 0.6-cores can be arrayed in a single master block, use of this core size allows increased throughput and decreased cost.
(Arch Pathol Lab Med. 2010;134:613-619)
Cellular proteins (Health aspects)
Cellular proteins (Research)
Growth factor receptors (Health aspects)
Growth factor receptors (Research)
Tissue microarrays (Usage)
Anagnostou, Valsamo K.
Lowery, Frank J.
Syrigos, Konstantinos N.
Cagle, Philip T.
Rimm, David L.
|Publication:||Name: Archives of Pathology & Laboratory Medicine Publisher: College of American Pathologists Audience: Academic; Professional Format: Magazine/Journal Subject: Health Copyright: COPYRIGHT 2010 College of American Pathologists ISSN: 1543-2165|
|Issue:||Date: April, 2010 Source Volume: 134 Source Issue: 4|
|Topic:||Event Code: 310 Science & research|
|Product:||Product Code: 2831720 Cellular Protein NAICS Code: 325413 In-Vitro Diagnostic Substance Manufacturing|
|Geographic:||Geographic Scope: United States Geographic Code: 1USA United States|
Tissue microarray (TMA) technology is a high-throughput research
tool for assessing protein expression levels in large cohorts and allows
for consecutive testing of multiple diagnostic and prognostic
biomarkers. (1,2) Typically, small-core (0.6-1.5 mm in diameter)
"biopsies" of standard formalin-fixed, paraffin-embedded
tissue in a histologic block are arrayed into a recipient
"master" block with minimal damage to the original blocks.2
Tissue from hundreds of specimens can thus be represented on a single
paraffin block, which allows maximization of tissue resources as well as
the processing of all cases in an identical manner. The number of
histospots on a single slide depends on the array design but is
typically 600 to 1000 for cores 0.6 mm in diameter (although arrays with
more than 800 cores are technically challenging). Larger cores, 1 mm and
1.5 mm in diameter, allow a maximum of only 200 or 100 cores,
respectively. Using spots 0.6 mm in diameter--at a spacing of 0.8
mm--allows for an area of observation of 0.282 mm, (2) which corresponds
to 2 to 3 high-power fields, conventionally sufficient for morphologic
tissue evaluation. (3) Tissue microarray construction that incorporates
spots 0.6 mm in diameter allows for simultaneous cohort analysis, while
minimizing the amount of reagents required and the cost of the assay.
Moreover, the same block can undergo several needle biopsies, 200 to 300
or more, depending on the tumor size in the original block; this reduces
the amount of tissue required for the particular study, thus preserving
the remaining tissue for future research or for diagnostic needs. (4)
The main criticism of the TMA technology concerns representative tissue sampling in heterogeneous lesions. Many studies of representativeness have compared TMAs to whole sections, (5,6) and selection of 1 to 10 cores per tumor sample has been used, depending on the biomarker heterogeneity and the tissue tested, (7) but few, if any, have compared the difference between core sizes. Moreover, no studies have performed the comparison in a quantitative manner. This is a relevant question because it ultimately determines the number of assays that can be done on a given tissue sample block and the maximum number of cores per block, which has implications for throughput. Here we examined the potential for using cores 0.6 mm in diameter as a standard for core sampling, as compared to the 1-mm and 1.5-mm alternatives in breast and lung cancer and melanoma.
MATERIALS AND METHODS
Cohorts and Tissue Microarrays
All tissue specimens were analyzed in a TMA format: representative tumor areas were obtained from formalin-fixed, paraffin-embedded specimens of primary breast (Yale University/New Haven Hospital, New Haven, Connecticut) and lung (Methodist Hospital, Houston, Texas) tumors and cores 1 mm and 1.5 mm in diameter from each tumor block were arrayed in recipient blocks for the lung and breast cancer TMA, respectively. The Yale University breast cancer cohort consisted of 36 primary breast cancer specimens, obtained from the archives of the Pathology Department of Yale University, and the Methodist Hospital lung cancer cohort consisted of 55 primary non-small cell lung cancer specimens obtained from the Pathology Department of Methodist Hospital. The study was approved by the institutional review boards of both centers. Tissue micro-arrays with 45 cases of 1-mm melanoma (2-fold redundancy) and 72 cores of 1.5-mm lung cancer were purchased from Cybrdi, Inc (Rockville, Maryland).
A modified immunofluorescent protocol was used, described in detail elsewhere.8 In brief, the arrays were deparaffinized with xylene, rehydrated, and antigen-retrieved by pressure cooking for 15 minutes in citrate buffer (pH = 6) for all primary antibodies but those for epidermal growth factor receptor (EGFR), for which proteinase K (Dako, Carpinteria, California) enzymatic digestion was performed instead. Slides were preincubated with 0.3% bovine serum albumin (BSA) in 0.05 M Trisbuffered saline with Tween-20 (TBST, pH = 8) for 30 minutes at room temperature.
Breast cancer slides were then incubated overnight with a cocktail of either the human epidermal receptor 2 (HER2/neu) primary antibody diluted 1:500 (rabbit polyclonal; Dako) and a mouse monoclonal anti-human cytokeratin antibody (clone AE1/AE3, M3515; Dako) diluted 1:100 in BSA/TBST or the estrogen receptor (ER) a primary antibody diluted 1:50 (mouse monoclonal, clone 1D5; Dako) and a wide-spectrum rabbit anti-cow cytokeratin antibody (Z0622; Dako) diluted 1:100 in BSA/ TBST. The Methodist Hospital lung cancer slides were incubated overnight with a cocktail of either the EGFR primary antibody diluted 1:500 (mouse monoclonal, clone 31G7; Zymed Laboratories, San Francisco, California) or the signal transducer and activator of transcription 3 (STAT3) primary antibody diluted 1:200 (mouse monoclonal, clone 124H6; Cell Signaling Technology Inc, Danvers, Massachusetts) and a wide-spectrum rabbit anti-cow cytokeratin antibody (Z0622; Dako) diluted 1:100 in BSA/TBST. The Cybrdi lung cancer TMAs were incubated overnight with a cocktail of either the mammalian target of rapamycin (mTOR) primary antibody diluted 1:200 (rabbit monoclonal, clone 7C10; Cell Signaling Technology) and a mouse monoclonal anti-human cytokeratin antibody (clone AE1/AE3, M3515; Dako) diluted 1:100 in BSA/TBST or the phospho-p70 S6 kinase (pS6K) (Thr389) primary antibody diluted 1:200 (mouse monoclonal, clone 1A5; Cell Signaling Technology) and a wide-spectrum rabbit anti-cow cytokeratin antibody (Z0622; Dako) diluted 1:100 in BSA/TBST. For the melanoma cohort, a mouse monoclonal S100 antibody (Z0311; Dako) and a rabbit polyclonal S100 antibody (15E2E2; BioGenex, San Ramon, California), both diluted 1:100 in BSA/TBST, were used instead of the mouse and rabbit cytokeratin, respectively.
This process was followed by a 1-hour incubation with Alexa 546-conjugated goat anti-mouse secondary antibody (A11003; Molecular Probes, Eugene, Oregon) diluted 1:100 in rabbit EnVision reagent (K4003; Dako) for mTOR and HER2/neu and incubation with Alexa 546-conjugated goat anti-rabbit secondary antibody (A11010; Molecular Probes) and diluted 1:100 in mouse EnVision reagent (K4001; Dako) for EGFR, ER, STAT3, and pS6K. Cyanine 5 directly conjugated to tyramide (FP1117, Perkin-Elmer, Boston, Massachusetts) at a 1:50 dilution was used as the fluorescent chromagen for target detection. ProLong mounting medium (ProLong Gold, P36931; Molecular Probes) containing 4', 6-diamidino-2-phenylindole was used to identify tissue nuclei. Serial sections of a smaller TMA consisting of 42 lung cancer specimens (control array) were stained aside all cohorts to confirm assay reproducibility. A431 cells were used as positive controls for EGFR and mTOR; MCF-7 cells as positive control for ER; HER2/neu-transfected Chinese hamster ovary cells as positive controls for HER2/neu; and breast cancer tissue as positive control for STAT3, as indicated by the manufacturers. Negative-control sections, in which the primary antibodies were omitted, were used for each immunostaining run.
Automated Quantitative Analysis
Automated quantitative analysis (now trademarked by HistoRx, New Haven, Connecticut, as AQUA) allows exact measurement of protein concentration within subcellular compartments, as described in detail elsewhere. (9) In brief, a series of high-resolution monochromatic images were captured by the PM-2000 microscope (HistoRx). For each histospot, in-focus and out-of-focus images were obtained by using the signal from the 4',6diamidino-2-phenylindole, cytokeratin-Alexa 546, and target protein (ER, HER2/neu, mTOR, pS6K, STAT3, EGFR)-cyanine 5 channel. The target protein was measured by using a channel with emission maxima above 620 nm, so as to minimize tissue autofluorescence. Epithelial tumors were distinguished from stromal and nonstromal elements by creating an epithelial tumor "mask" from the cytokeratin signal, while melanomas were identified by the S100 protein signal. This created a binary mask (each pixel being either "on" or "off") on the basis of an intensity threshold set by visual inspection of histospots. AQUA score of the target protein in each subcellular compartment was calculated by dividing the target compartment pixel intensities by the area of the compartment within which they were measured. AQUA scores were normalized to the exposure time and bit depth at which the images were captured. This strategy allowed for the direct comparison of scores collected at different exposure times. Specimens with less that 5% tumor area per spot were excluded.
Acquisition Algorithm for 1-mm and 1.5-mm Histospots
We used the PM-2000 epifluorescence microscopy platform and the SpotGrabber software (HistoRx) to identify, tag, and assign grid coordinates to TMA cores within a selected region in preparation for standard high-resolution (X20) acquisitions (Figure 1). Spots 1 mm and 1.5 mm in diameter do not entirely fit in the X 20 field of view and 1 random 0.6-mm field that corresponds to the X 20 objective field of view was acquired per histospot (left arm of acquisition algorithm; Figure 1). Alternatively, spots 1 mm and 1.5 mm in diameter were given a set of coordinates delimiting the perimeter of the histospot by Section-Grabber (HistoRx). Selections illustrated in the right arm of the algorithm (Figure 1) were automatically subdivided into a grid of 6 to 12 fields (depending on core size) and the PM-2000 microscope was programmed to take pictures of adjacent fields that together cover the entire area of the histospot. A graphic representation of automated quantitative analysis for mTOR in 0.6-mm and 1.5-mm lung cancer fields is shown in Figure 2. Average AQUA scores were calculated for each 1-mm and 1.5-mm core and compared to respective scores of 1 random 0.6-mm field.
The Pearson correlation coefficient (R) was used to assess the correlation between AQUA scores from the same cores on serial cuts of the control array; an R2 greater than 0.8 was indicative of excellent intra-array reproducibility. The association of mTOR, pS6K, EGFR, and STAT3 AQUA scores between 0.6 mm and 1 mm acquisitions and of HER2/neu,ER,mTOR, and pS6K between 0.6 mm and 1.5 mm acquisitions was also analyzed by using linear regression analysis and the Pearson correlation coefficient. The association between mTOR and pS6K AQUA scores was analyzed using Spearman rank test. All P values were based on 2-sided testing and differences were considered significant at P < .05. All statistical analyses were done by using the SPSS software program (version 13.0 for Windows; SPSS Inc, Chicago, Illinois).
[FIGURE 1 OMITTED]
We assessed the expression of a series of biomarkers (ER, HER2/neu, EGFR, STAT3, mTOR, and phospho-p70 S6K) by automated quantitative analysis in the standard 0.6-mm format (that corresponds to a conventional X20 high-resolution acquisition), as compared to 1-mm and 1.5-mm acquisitions. These markers were chosen on the basis of their common application in diagnostics and/or their heterogeneous expression; analysis of 2 well-established prognostic and predictive markers in breast cancer (ER and HER2/ neu) was also performed.
[FIGURE 2 OMITTED]
[FIGURE 3 OMITTED]
First, we assessed spot-size effect for ER and HER2/neu in breast cancer specimens by measuring 32 and 26 samples, respectively, consisting of 1.5-mm spots. Quantitative analysis of 1 random 0.6-mm field from the center of each 1.5-mm histospot in the breast cancer TMA yielded AQUA scores that ranged from 12.71 to 1281 (mean, 213.2 [+ or -] 49) for ER and 3 to 347.6 (mean, 30.64 [+ or -] 13.94) for HER2/ neu. The same range was observed when 6 to 12 fields per core were captured to reconstitute the full-size 1.5-mm spots. AQUA scores did not significantly vary among fields; therefore, average values for ER and HER2/ neu were calculated and treated as independent continuous variables (range, 14.14--863.15; mean, 194.4 [+ or -] 43.92 for ER; range, 3.71-246.8; mean, 23.5 [+ or -] 10.21 for HER2/neu). Measurement of a single random 0.6-mm spot was comparable to analysis of the whole 1.5-mm spot, with a Pearson R coefficient of 0.94, P < .001 for HER2/neu and R = 0.92, P < .001 for ER (Figure 3, A and B).
A lung cancer TMA consisting of seventy-two 1.5-mm spots (Cybrdi) was used to assess the 1.5-mm spot-size effect in mTOR and pS6K measurements. Expression of mTOR varied from 7.14 to 98.03 (mean, 28.59 [+ or -] 1.9) for 0.6mm spots and from 8.56 to 94.89 (mean, 25.08 [+ or -] 1.7) for 1.5-mm histospots. pS6K AQUA scores ranged from 50.55 to 635.98 (mean, 183.56 [+ or -] 15.25) and from 47.93 to 455.99 (mean, 155.08 [+ or -] 11.46) for 0.6-mm and 1.5-mm acquisitions, respectively. A single 0.6-mm field from the center of the spot detected the same amount of mTOR and pS6K protein (Pearson R = 0.96, P < .001 and R = 0.98, P < .001 for mTOR and pS6K, respectively; Figure 3, C and D) as that detected in large-core analysis. pS6K is the immediate downstream effector of mTOR, and previous studies (10) have shown a high correlation between mTOR and pS6K.
We found that using 0.6-mm fields from 1.5-mm histospots or using the average of the whole 1.5-mm histospot did not significantly change the correlation between mTOR and pS6K (Spearman [rho] = 0.36, = .004 and [rho] = 0.34, P = .004 for 0.6-mm and 1.5-mm acquisitions, respectively).
[FIGURE 4 OMITTED]
Assessment of mTOR and pS6K was also performed for melanoma, this time by comparing spot sizes of 0.6 mm and 1 mm. AQUA scores ranged from 7.61 to 148.43 (mean, 40.64 [+ or -] 4.95) for the 0.6-mm histospots for mTOR, as compared to 7.77 to 102.31 (mean, 30 [+ or -] 3.37, Pearson R = 0.92 P < .001; Figure 4, A) for 1.0-mm spots. For pS6K, AQUA scores ranged from 76.15 to 377.01 (mean, 172.53 [+ or -] 8.69) for 0.6-mm spots as compared to 65.35 to 350.12 (mean, 167.61 [+ or -] 8.87, Pearson R = 0.9, P < .001; Figure 4, B) for 1-mm spots. Interestingly, results for mTOR and pS6K were more strongly correlated in the 0.6-mm setting than they were with 1-mm spots (Spearman p = 0.62, P < .001 and p = 0.48, P < .001 for 0.6-mm and 1-mm histospots, respectively), although both were highly significant.
We next assessed the effect of 0.6-mm acquisitions versus 1-mm acquisitions in lung cancer. Epidermal growth factor receptor and STAT3 were measured by using 54 and 41 histospots from the Methodist Hospital lung cancer cohort arrayed with the 1-mm-core format. AQUA scores corresponding to virtual 0.6-mm spots, created as described above, ranged from 9.34 to 225.72 for EGFR (mean, 56.69 [+ or -] 8.02) and from 42.62 to 679.56 for STAT3 (mean, 21.63 [+ or -] 21). Average AQUA scores (representing 6-12 fields to cover the whole spot) for the 1-mm spots ranged from 6.2 to 225.01 for EGFR (mean, 45.5 [+ or -] 6.78) and 16.07 to 538.14 for STAT3 (mean, 129.79 [+ or -] 17.7). A highly significant correlation of AQUA scores was found between 0.6-mm and 1-mm acquisitions (Pearson R = 0.98, P < .001 and R = 0.87, P = .005 for EGFR and STAT3, respectively; Figure 4, C and D). Evaluation of the interarray reproducibility did not reveal significant differences between serial sections of the control array run aside each TMA (Pearson R > 0.9 for all runs, P < .001).
Although many studies11 have focused on the number of TMA spots required to represent a conventional histologic section, the ideal spot size for TMAs has largely been a controversial topic, driven by subjective assessment and opinion. Here, we use the objective AQUA-based method to demonstrate that TMAs with 0.6-mm cores are equivalent to the 1-mm and 1.5-mm histospots with respect to estimation of protein concentration. We assessed the expression of ER, HER2/ neu, EGFR, STAT3, mTOR, and pS6K in 3 common tumors (breast and lung cancer and melanoma) to avoid bias associated with a single tumor type. Estrogen receptor and HER2/ neu were chosen because they are routinely used on diagnostic specimens; EGFR and STAT3 were selected on the basis of their homogeneous and heterogeneous expression, respectively, in lung cancer (V.K.A., unpublished data, 2009). Finally, mTOR and pS6K were assessed in 2 tumor types, since they have an affector/effector relationship and are emerging companion diagnostic biomarkers. We conclude that the expression of each of these biomarkers was the same among all core sizes for all tumor types studied.
Torhorst et al (6) have shown that tens of thousands of TMA sections can be generated from a single master 3-mm-thick block containing a 10 3 10-mm tumor area.
Given that up to one thousand 0.6-mm cores can be evaluated on a single microscope slide, use of 0.6-mm cores increases the speed of analysis of large cohorts and facilitates the interpretation of the results, therefore allowing increased throughput and decreased cost. Furthermore, if the availability of tissue is a rate-limiting step for biomarker studies, 0.6-mm-core TMA analysis significantly reduces the amount of archival material required for a particular study, thus preserving remaining tissue for other purposes. (12)
Sampling of heterogeneous tumors is the major limitation of the TMA technique, and the ability of TMAs to reproduce large sections has been largely questioned with reports on significant tumor heterogeneity reaching 81% for ER expression in breast cancer. (7,13) Several studies have now shown that TMA analysis provides the same clinicopathologic associations reported on large tissue specimens (2,6,8,14-16) and that 0.6-mm-core TMAs with 2-fold redundancy provide an accurate method for biomarker analysis in large archival cohorts. (17) In this study, we did not investigate the differences in associations among biomarker expression and clinical outcome between 0.6mm and larger cores, although we have used this core size extensively in previous outcome studies. (11)
A common argument for the use of larger spots is that 0.6-mm specimens are not interpretable because some critical features are of insufficient density. For example, features like blood vessels, lymphatic vessels, or stem cells may occur at insufficient frequency for assessment in a smaller core. However, recent studies have shown that features traditionally assessed in larger areas, like grade and Ki-67 index, can be reliably measured in smaller cores, (18) even in the setting of heterogeneous diseases like Hodgkin lymphoma. (19) This work does not attempt to assess these architectural features, but rather focuses on protein expression. While it is possible that larger cores are better for counting architecturally rare events, we are not aware of studies specifically addressing the issue. Ongoing work in our laboratory suggests that no core size (0.6-1.5 mm) will be sufficient for 1 such architectural feature (20) and, in that case, only a full-section analysis will suffice. The fact that we have not assessed vessels or similar architectural features as a function of spot size represents a limitation of this work. However, we have not attempted this because we believe that those features question the value of the TMA and require spot versus whole-slide comparisons, as opposed to spot-size comparisons.
In summary, we have shown that measurement of a single random 0.6-mm spot is comparable to analysis of whole 1-mm or 1.5-mm spots for a series of biomarkers in lung and breast cancer and melanoma. Our findings suggest that TMAs incorporating 0.6-mm cores may be used as a gold standard for large biomarker studies in which the key measurement is biomarker protein expression. Using this standard could marginally increase available tissue resources, but more importantly, it decreases slide-to-slide variability and reduces the cost per assay, without sacrificing accuracy. This observation may be of value as the US Food and Drug Administration and other regulatory agencies begin more aggressive regulation of companion diagnostics.
(1.) Rimm DL, Camp RL, Charette LA, Olsen DA, Provost E. Amplification of tissue by construction of tissue microarrays. Exp Mol Pathol. 2001;70(3): 255-264.
(2.) Kononen J, Bubendorf L, Kallioniemi A, et al. Tissue microarrays for high- throughput molecular profiling of tumor specimens. Nat Med. 1998;4(7):844-847.
(3.) Rimm DL, Camp RL, Charette LA, Costa J, Olsen DA, Reiss M. Tissue microarray: a new technology for amplification of tissue resources. Cancer J. 2001;7(1):24-31.
(4.) Giltnane JM, Rimm DL. Technology insight: identification of biomarkers with tissue microarray technology. Nat Clin Pract. 2004;1(2):104-111.
(5.) Iakovlev VV, Pintilie M, Morrison A, Fyles AW, Hill RP, Hedley DW. Effect of distributional heterogeneity on the analysis of tumor hypoxia based on carbonic anhydrase IX. Lab Invest. 2007;87(12):1206-1217.
(6.) Torhorst J, Bucher C, Kononen J, et al. Tissue microarrays for rapid linking of molecular changes to clinical endpoints. Am J Pathol. 2001;159(6):2249-2256.
(7.) Chung GG, Zerkowski MP, Ghosh S, Camp RL, Rimm DL. Quantitative analysis of estrogen receptor heterogeneity in breast cancer. Lab Invest. 2007; 87(7):662-669.
(8.) McCabe A, Dolled-Filhart M, Camp RL, Rimm DL. Automated quantitative analysis (AQUA) of in situ protein expression, antibody concentration, and prognosis. J Nat Cancer Inst. 2005;97(24):1808-1815.
(9.) Camp RL, Chung GG, Rimm DL. Automated subcellular localization and quantification of protein expression in tissue microarrays. Nat Med. 2002;8(11): 1323-1327.
(10.) Sahin F, Kannangai R, Adegbola O, Wang J, Su G, Torbenson M. mTOR and P70 S6 kinase expression in primary liver neoplasms. Clin Cancer Res. 2004; 10(24):8421-8425.
(11.) Camp RL, Neumeister V, Rimm DL. A decade of tissue microarrays: progress in the discovery and validation of cancer biomarkers. J Clin Oncol. 2008;26(34):5630-5637.
(12.) Mills SE, Fechner RE, Frierson HF, et al. Guardians of the wax and the patient. Am J Clin Pathol. 1995;104(4):365-367.
(13.) Layfield LJ, Saria E, Mooney EE, Liu K, Dodge RR. Tissue heterogeneity of immunohistochemically detected estrogen receptor: implications for image analysis quantification. Am J Clin Pathol. 1998;110(6):758-764.
(14.) Moch H, Schraml P, Bubendorf L, et al. High-throughput tissue microarray analysis to evaluate genes uncovered by cDNA microarray screening in renal cell carcinoma. Am J Pathol. 1999;154(4):981-986.
(15.) Schraml P, Kononen J, Bubendorf L, et al. Tissue microarrays for gene amplification surveys in many different tumor types. Clin Cancer Res. 1999;5(8): 1966-1975.
(16.) Kyndi M, Sorensen FB, Knudsen H, et al. Tissue microarrays compared with whole sections and biochemical analyses: a subgroup analysis of DBCG 82 b&c. Acta Oncol. 2008;47(4):591-599.
(17.) Camp RL, Charette LA, Rimm DL. Validation of tissue microarray technology in breast carcinoma. Lab Invest. 2000;80(12):1943-1949.
(18.) Nocito A, Bubendorf L, Tinner EM, et al. Microarrays of bladder cancer tissue are highly representative of proliferation index and histological grade. J Pathol. 2001;194(3):349-357.
(19.) Garcia JF, Camacho Fl, Morente M, et al. Hodgkin and Reed-Sternberg cells harbor alterations in the major tumor suppressor pathways and cell-cycle checkpoints: analyses using tissue microarrays. Blood. 2003;101(2):681-689.
(20.) Robinson BD, YL, Ingram C, Rohan T, Condeelis J, Jones JG. Quantitation of TMEM (tumor microenvironment of metastasis) in human primary breast cancer samples is predictive of long-term metastatic potential. United States and Canadian Academy of Pathology Annual Meeting; 2008. Abstract 1504.
Valsamo K. Anagnostou, MD; Frank J. Lowery; Konstantinos N. Syrigos, MD, PhD; Philip T. Cagle, MD; David L. Rimm, MD, PhD
Accepted for publication June 12, 2009.
From the Department of Pathology, Yale University School of Medicine, New Haven, Connecticut (Drs Anagnostou, Syrigos, and Rimm and Mr Lowery);and the Department of Pathology, The Methodist Hospital, Houston, Texas (Dr Cagle).
The authors have no relevant financial interest in the products or companies described in this article.
Reprints: David L. Rimm, MD, PhD, Department of Pathology, Yale University School of Medicine, 310 Cedar Street, , PO Box 208023, New Haven, CT 06520-8023 (e-mail: firstname.lastname@example.org).
|Gale Copyright:||Copyright 2010 Gale, Cengage Learning. All rights reserved.|