BACKGROUND: The tumour suppressor gene phosphatase and tensin homologue (PTEN) is considered a potential biomarker although results are inconsistent and there is debate as to the optimal method of reporting low PTEN expression/loss of function. Here we report results of a subgroup of patients from the AGITG/MAX trial PTEN analysis to explore the potential difference in concordance between pathologists assessment of IHC, and the Taqman copy number.
METHODS: Genomic DNA was extracted from FFPE tissue sections. The Taqman PTEN copy number assay (CNA) was performed using 5 ng DNA in quadruplicate PCR. The results are calculated as a ratio relative to a 2-copy control using the 2-∆∆Ct method (Rotorgene software), and multiplied by 2 to give the copy number. Loss of PTEN was defined as <1.5 copies, no loss was > 1.5 copies. IHC was performed on 59 tissue samples on tissue arrays and assessed by two blinded Pathologists. Two pathologist were compared directly for an IHC concordance rate, and then the scores were grouped and majority score used for IHC v Taqman CNA concordance rate.
RESULTS: Tissue samples on tissue arrays were analysed for PTEN staining by IHC. 3 tissues were found to have no residual tissue present on the array and 96% of tissue was from primary lesion. 18 of 56 (32%) had score of zero, ie loss of PTEN expression based on IHC expression. Concordance of IHC PTEN loss was however only 66% (37/56) when comparing two pathologists. Using majority score for IHC from the two pathologists, the concordance with Taqman CNA was 64%.

CONCLUSION: Inter-observer variability of PTEN IHC assessment as demonstrated by lack of concordance between pathologists highlights the limitation of this approach to PTEN assessment. There also appeared little evidence that PTEN loss based on IHC was indicative of copy number loss.