Quantitative analysis of estrogen receptor heterogeneity in breast cancer
Quantitative analysis of estrogen receptor heterogeneity in breast cancer
复制标题
DOI：
10.1038/labinvest.3700543
复制
发表时间：
2007-07-01
影响因子：
5
通讯作者：
Rimm, David L.
中科院分区：
文献类型：
--
作者：
Chung, Gina G.;Zerkowski, Maciej P.;Rimm, David L.
生物标志物的免疫组织化学分析（IHC）广泛用于肿瘤表征以及作为预后和预测措施。目前的单载玻片分析标准假定一个5 mM切片代表整个肿瘤。我们使用我们的自动图像分析技术（AQUA），使用改良的IHC技术和荧光团，比较原发性乳腺癌病例的多个组织块/载玻片中的雌激素受体（ER）表达，目的是量化切片内和组织块之间的肿瘤异质性。为了使我们的ER评分标准化，并允许载玻片与载玻片的比较，将具有已知ER评分的代表性乳腺癌病例的0.6 μ m组织点组装成“金标准阵列”（GSA），并将其邻近每个完整切片放置。总体而言，AQUA评分与病理学家评分和GSA评分的再现性之间存在良好的相关性（平均线性回归R值为0.8903）。然后分析了来自11例手术病例的29张载玻片，总计超过2000张AQUA图像。使用AQUA（> 10）和病理学家（> 10%）评分的标准二元分配为阳性，AQUA和病理学家评分（73%）之间以及来自相同病例的不同组织块的载玻片之间（75%）存在相当的一致性。然而，使用连续AQUA评分，AQUA和病理学家之间的一致性要低得多，同一病例不同组织块的载玻片之间的一致性仅为19%。在单个载玻片内，也存在分散模式的显著异质性，最明显的是AQUA评分最高的载玻片。总之，使用ER表达的定量测量，在81%的病例中发现了显著的块间异质性。这些结果很可能反映了由于缺乏免疫组织化学标准化和真实生物异质性而导致的实验室变异性。这也可能取决于所分析的生物标志物，并建议应进行进一步的研究，以确定这些发现如何影响临床决策过程。
Immunohistochemical analyses (IHC) of biomarkers are extensively used for tumor characterization and as prognostic and predictive measures. The current standard of single slide analysis assumes that one 5 mM section is representative of the entire tumor. We used our automated image analysis technology (AQUA) using a modified IHC technique with fluorophores to compare estrogen receptor (ER) expression in multiple blocks/slides from cases of primary breast cancer with the objective of quantifying tumor heterogeneity within sections and between blocks. To normalize our ER scores and allow slide-to-slide comparisons, 0.6 mu m histospots of representative breast cancer cases with known ER scores were assembled into a 'gold standard array' (GSA) and placed adjacently to each whole section. Overall, there was excellent correlation between AQUA scores and the pathologist's scores and reproducibility of GSA scores (mean linear regression R value 0.8903). Twenty-nine slides from 11 surgical cases were then analyzed totaling over 2000 AQUA images. Using standard binary assignments of AQUA (> 10) and pathologist's (> 10%) scores as being positive, there was fair concordancy between AQUA and pathologist scores (73%) and between slides from different blocks from the same cases (75%). However using continuous AQUA scores, agreement between AQUA and pathologist was far lower and between slides from different blocks from the same cases only 19%. Within individual slides there was also significant heterogeneity in a scattered pattern, most notably for slides with the highest AQUA scores. In sum, using a quantitative measure of ER expression, significant block-to-block heterogeneity was found in 81% of cases. These results most likely reflect both laboratory-based variability due to lack of standardization of immunohistochemistry and true biological heterogeneity. It is also likely to be dependent on the biomarker analyzed and suggests further studies should be carried out to determine how these findings may affect clinical decision-making processes.