Tissue Microarray Sampling Strategy for Prostate Cancer Biomarker Analysis Mark A. Rubin, M.D., Rodney Dunn, M.S., Myla Strawderman, M.S., and Kenneth J. Pienta, M.D. High-density tissue microarrays (TMA) are useful for profiling protein expression in a large number of samples but their use for clinical biomarker studies may be limited in heterogeneous tumors like prostate cancer. In this study, the optimization and validation of a tumor sampling strategy for a prostate cancer outcomes TMA is performed. Prostate cancer proliferation de- termined by Ki-67 immunohistochemistry was tested. Ten rep- licate measurements of proliferation using digital image analy- sis (CAS200, Bacus Labs, Lombard, IL, USA) were made on 10 regions of prostate cancer from a standard glass slide. Five matching tissue microarray sample cores (0.6 mm diameter) were sampled from each of the 10 regions in the parallel study. A bootstrap resampling analysis was used to statistically simu- late all possible permutations of TMA sample number per re- gion or sample. Statistical analysis compared TMA samples with Ki-67 expression in standard pathology immunohisto- chemistry slides. The optimal sampling for TMA cores was reached at 3 as fewer TMA samples significantly increased Ki-67 variability and a larger number did not significantly im- prove accuracy. To validate these results, a prostate cancer outcomes tissue microarray containing 10 replicate tumor samples from 88 cases was constructed. Similar to the initial study, 1 to 10 randomly selected cores were used to evaluate the Ki-67 expression for each case, computing the 90th per- centile of the expression from all samples used in each model. Using this value, a Cox proportional hazards analysis was per- formed to determine predictors of time until prostate-specific antigen (PSA) recurrence after radical prostatectomy for clini- cally localized prostate cancer. Examination of multiple models demonstrated that 4 cores was optimal. Using a model with 4 cores, a Cox regression model demonstrated that Ki-67 expres- sion, preoperative PSA, and surgical margin status predicted time to PSA recurrence with hazard ratios of 1.49 (95% con- fidence interval [CI] 1.01–2.20, p  0.047), 2.36 (95% CI 1.15–4.85, p  0.020), and 9.04 (95% CI 2.42–33.81, p  0.001), respectively. Models with 3 cores to determine Ki-67 expression were also found to predict outcome. In summary, 3 cores were required to optimally represent Ki-67 expression with respect to the standard tumor slide. Three to 4 cores gave the optimal predictive value in a prostate cancer outcomes ar- ray. Sampling strategies with fewer than 3 cores may not ac- curately represent tumor protein expression. Conversely, more than 4 cores will not add significant information. This prostate cancer outcomes array should be useful in evaluating other putative prostate cancer biomarkers. Key Words: Tissue microarray—Expression arrays— Biomarkers—Prostate cancer—PSA recurrence—Proliferation— Ki-67—Digital image analysis. Am J Surg Pathol 26(3): 312–319, 2002. High-density tissue microarrays (TMA) are useful for rapid analysis of molecular markers in large number of samples (“protein profiling”). 12 TMA technology in- creases by an order of magnitude the number of speci- mens available for analysis while causing minimal de- struction to the original donor material. The use of TMAs for clinical biomarker studies, however, may be limited in heterogeneous tumors like prostate carcinoma (PCA) because of inadequate tissue sampling. However, TMAs play an exceedingly important in a high-throughput para- digm that uses cDNA expression microarrays and TMAs to discover and then validate candidate biomarkers. 5 Therefore, the overall goal of the present study is to optimize sampling for the evaluation TMAs to identify PCA biomarkers. Biologic differences between tumor samples can be quantified by using biomarkers. A good biomarker adds independent information to standard prognosticators such as pathologic and clinical stage. 18 Ki-67 is a nuclear protein that is expressed in G1, S, G2, and M phases of the cell cycle and is not seen at rest (G0 phase). 6 The Ki-67 protein can be readily detected in formalin-fixed, paraffin-embedded tissue with the monoclonal antibody Mib-1. Ki-67 expression is most often defined as the number of positive staining nuclei per total nuclei, termed Ki-67 Labeling Index (Ki-67 LI). The prognostic significance of Ki-67 LI in PCA has been examined in several studies. 1,3,4,10,11,17 Significant associations have From the Departments of Pathology (M.A.R.), Urology (M.A.R., R.D., K.J.P.), Internal Medicine (K.J.P.), Division of Hematology/Oncology, Biostatistics (R.D., M.S.), and the University of Michigan Comprehensive Cancer Center (M.A.R., R.D., M.S., K.J.P.), Ann Arbor, Michigan, U.S.A. Supported by Specialized Program of Research Excellence for Pros- tate Cancer (SPORE) NCI Grant P50CA69568. Address correspondence and reprint requests to Mark A. Rubin, MD, University of Michigan, Department of Pathology, 1500 E. Medical Center Drive, Room 2G322/Box 0054, Ann Arbor, MI 48109-0054, U.S.A.; e-mail: marubin@umich.edu The American Journal of Surgical Pathology 26(3): 312–319, 2002 © 2002 Lippincott Williams & Wilkins, Inc., Philadelphia 312