Assay result variability during determination of mismatch repair deficiency status using immunohistochemistry: A transatlantic comparative study

Autor: Frederick L. Baehner, Kelly Handley, Richard Gray, Philip Quirke, Margarita Lopatin, Mark Lee, Gordon G A Hutchins, D. Kerr, Matthew T. Seymour, H. Yaziji
Rok vydání: 2011
Předmět:
Zdroj: Journal of Clinical Oncology. 29:403-403
ISSN: 1527-7755
0732-183X
Popis: 403 Background: Colorectal cancer patients with deficient mismatch repair (dMMR) have significantly fewer recurrences and may respond less well to chemotherapy. Immunohistochemical (IHC) determination of MMR status is therefore recommended to identify patients in whom adjuvant therapy is not indicated, but little is known regarding the variability of assay results. We aimed to define MMR IHC assay variability in formalin-fixed, paraffin-embedded (FFPE) material from the QUASAR-1 colorectal cancer trial contained within heterogeneity-prone tissue microarrays (TMAs) Methods: TMA sections of FFPE material from the QUASAR-1 trial were distributed to 2 independent laboratories (Leeds, UK and Vitro Molecular Laboratories [VML], FL, USA) for MMR IHC assays. Serial TMA sections were stained with MLH1/MSH2 using techniques blinded to the other laboratory. Each stained section was double-scored independently and results compared to determine inter-assay variability. Results: Matched MMR data were available for 1224 cases of MLH1 and 1223 cases of MSH2. Of these, loss of expression of MMR (dMMR) was reported in 160 cases (13.1%) by VML and 179 cases (14.6%) by Leeds. 140 (11.4%) were dMMR in both labs, 20 cases (1.6%) by VML alone and 39 (3.2%) by Leeds alone. Discordant dMMR status was observed in 56/166 (34%) dMMR cases identified with MLH1 and 10/34 (29%) cases identified with MSH2. Kappa coefficients for inter-assay agreement were 0.798 (95% CI = 0.748 - 0.848) for MMR status overall, 0.778 (95% CI = 0.722 - 0.835) for MLH1 and 0.823 (95% CI = 0.714 - 0.932) for MSH2. 85% (33/39) of discordant Leeds dMMR colon cancers were in the right colon where dMMR is more common, compared to 47% of VML discordant dMMR cases. Conclusions: Independent determination of MMR status by IHC on TMAs is associated with excellent inter-assay agreement. The reasons for MMR case discordance are under further investigation. These results further support routine MMR testing by IHC. No significant financial relationships to disclose.
Databáze: OpenAIRE
Externí odkaz: