mRNA profiling of a well-differentiated G1 pancreatic NET correlates with immunohistochemistry profile: a case report

Abstract
Background
Neuroendocrine neoplasms (NENs) are a complex group of tumours that occur in many organs. Routinely used IHC markers for NEN diagnosis include CgA, synaptophysin, Ki67 and CD56. These have limitations including lack of correlation to clinical outcomes and their presence in non-tumour tissue. Identification of additional markers and more quantitative analyses of tumour tissue has the potential to contribute to improved clinical outcomes. We used qRT-PCR to profile the expression levels of a panel of markers in tumour and matched non-tumour tissue from a patient with a G1 pancreatic neuroendocrine tumour. Differences in mRNA levels between tumour and non-tumour tissue were compared with IHC analyses of the same sample.

Case presentation
An elderly man presented with lower abdominal pain for 6 months. Histological analysis identified a low grade, well differentiated pancreatic endocrine neoplasm. Twenty-seven tumour markers for neuroendocrine status, proliferation, stem cell phenotype, angiogenesis, epithelial to mesenchymal transition, cell adhesion, differentiation and tumour suppression were selected from previous studies and mRNA levels of these markers were measured in tumour and adjacent non-tumour tissue sample using qRT-PCR. IHC was carried out on the same tissue to detect the corresponding marker proteins. Of the markers analysed, seven showed higher mRNA levels in tumour relative to non-tumour tissue while thirteen had lower expression in tumour relative to non-tumour tissue. Substantial differences in mRNA levels were a gain of CgA, CD56, β-catenin, CK20, PDX1 and p53 and loss of Ki67, PCAD, CK7, CD31, MENA, ECAD, EPCAM, CDX2 and CK6. Comparison of qRT-PCR data with IHC showed correlation between fifteen markers.

Conclusion
Our study is unique as it included matched controls that provided a comparative assessment for tumour tissue analysis, whereas many previous studies report tumour data only. Additionally, we utilised qRT-PCR, a relatively quantitative diagnostic tool for differential marker profiling, having the advantage of being reproducible, fast, cheap and accurate. qRT-PCR has the potential to improve the defining of tumour phenotypes and, in combination with IHC may have clinical utility towards improving tumour stratification or distinguishing tumour grades. The results need to be validated with different grades of NENs and related to clinical outcomes.