Skip to main content

Tissue microarrays compared with whole sections and biochemical analyses. A subgroup analysis of DBCG 82 b&c.

Kyndi M, Sørensen FB, Knudsen H, Overgaard M, Nielsen HM, Andersen J, Overgaard J

Publication: Acta Oncol (2008) • Vol 47 • Page 591-9PMID: 18465327View on PubMed

Purpose

The purpose of this paper was to determine if the use of tissue microarray cores, at different locations both within and between different paraffin blocks affects the reproducibility of estrogen receptor (ER), progesterone receptor (PR) and human epidermal growth factor receptor 2 (HER-2) case status determined by immunohistochemistry. Comparisons were also drawn between immunohistochemical analysis of whole sections and biochemical analysis of tumor homogenates.

Conclusion

Proportional agreements in receptor case status among TMA cores and corresponding whole sections were antigen-dependent although trends common to all three antigens were observed. Proportional agreement in cases status was greater when two TMA cores from the center of the tumor were compared as opposed to the periphery, and relatively unaffected by the addition of up to four TMA cores. Although proportional agreements for case status ranged from almost perfect to substantial for all three antigens when the number and location of TMA cores were compared to one another or to a whole section, Kappa values (k) were still lower than those observed when two whole sections were compared. The percent agreement for case status between a TMA core and a corresponding whole section was reported as 62% and 67% for ER and PR, respectively. Use of TMA cores for immunohistochemistry was further validated by comparing resultant ER and PR case status with biochemical determinations of free cytolosolic ER and PR in tumor homogenates, revealing an 85% agreement for ER status (k=0.64) and a 82% agreement for PR (k=0.62).

Studies (2)

Purpose

Proportional agreement (represented by the Kappa value, k) between receptor case status was determined by immunohistochemical analysis of (1) a 1.0 mm TMA core and a corresponding whole section, (2) 1.0 mm TMA cores taken from the center or periphery of the same paraffin block, (3) 1.0 mm cores taken from the same location of two different paraffin blocks, and (4) whole sections from two different paraffin blocks. A total of 108 paraffin blocks from 27 patients were examined.

Summary of Findings

Proportional agreements in receptor case status among TMA cores and corresponding whole sections were antigen-dependent. When a corresponding TMA core and whole section were compared almost perfect agreements were reported for ER (k=0.81), and substantial agreements for PR (k=0.70) and HER-2 (k=0.67). Similarly, comparisons between TMA cores punched at the core or periphery of a paraffin block revealed almost perfect agreement for ER (k=0.83) and PR (k=0.80) case status, and substantial agreement for HER-2 (k=0.73). When TMA cores taken from two different paraffin blocks were compared, the proportional agreement was higher among cores taken from the center as opposed to the periphery for all three antigens, ER (k=0.89 versus 0.77), PR (k=0.65 versus 0.54), and HER-2 (k=0.78 versus 0.69). Agreements among TMA cores did not increase when all four TMA cores were compared for ER (k=0.83), PR (k=0.67), or HER-2 (k=0.73). Importantly, agreements among TMA cores were lower than agreements between whole sections from two different paraffin blocks, which approached perfect for all three antigens investigated (ER, k=0.92; PR, k=1.00; HER-2, k=0.95). An additional source of variability included discrepancies among observers. Although the Kappa values reported above represent an average, observer values were reported separately in the manuscript and interobserver agreement was determined to be substantial for both TMA cores (k=0.75) and whole sections (k=0.65). The percent agreement between a TMA core and a corresponding whole section was reported as 62% and 67% for ER and PR respectively, although the case status criteria differed between proportional agreement analysis (for which a positive case was defined as nuclear staining in greater than 10% of invasive tumor cells) and percent agreement (for which a positive case was defined as nuclear staining in more than 10 or 50 invasive tumor cells in a TMA core or whole section, respectively).
  • Biospecimens

    TissueBreast
  • Preservation Types

    Formalin
  • Diagnoses

    NeoplasticCarcinoma
  • Platforms

    ProteinImmunohistochemistry
    ProteinTissue microarray
  • Pre-analytical Factors

    ClassificationPre-Analytical FactorValue(s)
    Biospecimen Aliquots and ComponentsAliquot size/volume
    Single paraffin block
    Two paraffin blocks
    Biospecimen Aliquots and ComponentsBiospecimen heterogeneity
    Biospecimen core
    Biospecimen periphery
    Biospecimen Aliquots and ComponentsType of slide
    Tissue microarray core
    Whole tissue section
    Immunohistochemistry SpecificTargeted peptide/protein
    ER
    HER-2
    PR