Validation of tissue microarray technology in vulvar cancer.
Fons G, van der Velden J, Burger M, ten Kate F
Purpose
The purpose of this paper was to compare immunohistochemical (IHC) staining in tissue microarrays (TMAs) versus whole section slides of vulvar cancer specimens.
Conclusion
Between 83% and 88% of cores were assessable and 65-72% of cases had 3 assessable cores. When a two class scoring system was used, concordant Ki-67, p16INK4, cyclin D1, epidermal growth factor receptor (EGFR), caspase-3 and cycloxygenase-2 (COX-2) staining was observed between the whole section and the TMA in 92%, 92%, 86%, 83%, 78%, and 89% of cases, respectively, but use of a three class scoring system led to a decrease in concordance between whole sections and TMAs. Generally, discordant COX-2 and EGFR staining between the whole section and the TMA was due to negative staining on the whole section slide and positive on the TMA, but discordant Ki-67 and caspase-3 staining generally arose from negative staining on the TMA and positive staining on the whole section. There was no pattern to the discordant results obtained between whole sections and TMA for cyclin D1 and p16INK4. Use of an alternate COX-2 scoring system that did not take into account staining intensity decreased concordance between the TMA and the whole section from 89% to 84%.
Studies (1)
Purpose
The purpose of this study was to compare COX-2, caspase-3, p16INK4, cyclin D1 and Ki-67 IHC staining in TMAs versus whole section slides of vulvar cancer specimens. The type of fixative used was not specified. The array contained three 0.6 mm cores for each of the 40 cases. Scoring was based on a 3 class system or as positive and negative. Only the system used for COX-2 and EGFR took into account stain intensity. An alternate scoring system for COX-2 that did not take into account intensity was also used. Results from the cores were added to get a tumor score. When there were discrepant results among cores from the same tumor, the staining category of the majority was used or if there were two cores, the higher category was used. Only cases with at least 2 assessable scores were included in the analysis.
Summary of Findings
Between 83% and 88% of cores were assessable and 65-72% of cases had 3 assessable cores. 3 slides were non-assessable due to inadequate staining and one case had discordant p16INK4 staining between the only 2 assessable cores. When a two class scoring system was used, concordant Ki-67, p16INK4, cyclin D1, EGFR, caspase-3 and COX-2 staining was observed between the whole section and the TMA in 92%, 92%, 86%, 83%, 78%, and 89% of cases, respectively. Use of a three class scoring system led to a decrease in concordance between whole sections and TMAs where concordant Ki-67, p16INK4, cyclin D1, EGFR, caspase-3 and COX-2 staining was observed between the whole section and the TMA in 73%, 89%, 80%, 68%, 64%, and 80% of cases, respectively. Generally, discordant COX-2 and EGFR staining between the whole section and the TMA was due to negative staining on the whole section slide and positive staining on the TMA, but discordant Ki-67 and caspase-3 staining generally arose from negative scores on the TMA and positive scores on the whole section. There was no pattern to the discordant results obtained between whole sections and TMA for cyclin D1 and p16INK4. Use of an alternate COX-2 scoring system that did not take into account staining intensity decreased concordance between the TMA and the whole section from 89% to 84%.
Biospecimens
TissueVulvaPreservation Types
Other PreservativeDiagnoses
NeoplasticCarcinomaPlatforms
ProteinImmunohistochemistryProteinTissue microarrayPre-analytical Factors
Classification Pre-Analytical Factor Value(s) Biospecimen Aliquots and Components Type of slide Tissue microarrayWhole tissue sectionTissue microarray Specific Data handling 2-class scoring3-class scoringImmunohistochemistry Specific Targeted peptide/protein COX-2Caspase-3Cyclin D1EGFRKi-67p16INK4