Vol 57, Supp. A (2006)
Original paper
Published online: 2006-09-25

open access

Page views 590
Article views/downloads 1464
Get Citation

Connect on Social Media

Connect on Social Media

Percentage of neoplastic cells in papillary thyroid carcinoma: implications for DNA microarray studies of gene expression profile

Ewa Chmielik, Michał Jarząb, Dariusz Lange

Abstract

Studies of gene expression profile using DNA microarray technology are usually performed using either tumor-derived sample material or isolated neoplastic cells obtained through microdissection. The scope of information about neoplastic transformation gained from studying profile of gene expression in microdissected samples would be much wider but collection of sufficient amounts of intact RNA is very difficult. A compromise could be reached by relating gene expression profile to percentage of neoplastic cells in the investigated tissue sample. The ratio of neoplastic cells in the investigated sample of papillary thyroid cancer was assessed through evaluation of approximated count of cell number in 10-18 examined image fields. This information was related to gene expression profiles obtained from DNA microarrays. The study involved 40 cases of papillary thyroid cancer; the percentage of PTC cells varied between 20 and 95% and only in half of the cases exceeded 75%. Correct differentiation of tumor and normal sample by means of gene expression profile was possible only when the percentage of tumor cells exceeded 25-30%. Seventeen genes showing the best correlation with the tumor cell numbers were selected and their classification potential was evaluated.
Strength of information derived from gene expression profile studies by DNA microarrays in papillary thyroid cancer cells is very reliable and permits distinguishing correctly between normal and neoplastic tissues even when the percentage of cancer cells does not exceed 25-50%. However, the differentiation potential of gene expression profile is not markedly improved by selection of genes showing best correlation with the number of tumor cells.

Article available in PDF format

View PDF (Polish) Download PDF file