Corsi brevi - Siapec

PATOLOGIA MOLECOLARE
Fig. 1.
The correct and early diagnosis of pancreatic cancers is another
important challenge that is encountered with this type
of tumor, also because its typical symptoms (abdominal pain,
unexpected weight loss, jaundice) are rather unspecific and
can point to a variety of different GI tract problems, with
chronic pancreatitis, that is a persistent inflammatory disease,
being the clinically most relevant differential diagnosis
for pancreatic cancer. In fact, in the course of chronic pancreatitis
inflammatory tumors may develop in the pancreas
causing the same signs and symptoms as malignant pancreatic
tumors. Malignant and inflammatory tumors are frequently
indistinguishable by conventional imaging modalities such
as computed tomography (CT), abdominal (US) or endoscopic
ultrasound (EUS), thus requiring cytological analysis
of cells obtained by US-, CT- or EUS-guided fine needle aspiration
biopsy (FNAB). However, the reliability of the
largely morphology-based cytological analyses of fine needle
aspirates of pancreatic tumors remains unsatisfactory with a
diagnostic accuracy between 60% and 80% 11-15 . Well-differentiated
carcinomas may escape recognition because of the
minimal cytological atypia they display. Conversely, chronic
pancreatitis may give rise to atypical cells that can be mistaken
for neoplastic cells. For both, malignant and benign tumors,
diagnosis is extremely difficult when intact cells in the
aspirate are rare or completely missing.
DNA arrays with their potential to assess the transcriptional
activity of many genes simultaneously are ideal tools for diagnostic
approach relying on the analysis of multiple genetic
markers rather than morphological evaluation of biopsy material.
In collaboration with Prof. Thomas M. Gress (University of
Ulm, Germany) we have tried to develop a specialized cDNA
array designed for the differential diagnosis of pancreatic tumors
based on expression profiling of fine needle aspiration
biopsies.
Diagnostic array was constructed to only contain genes with
diagnostic and/or prognostic potential for the classification
of pancreatic tissues, augmented with control features to
allow for precise grid alignment and robust normalization. In
the present study, we used residual material from biopsy
needles for the analysis of the FNAB samples to ensure
complete identity of the material used for cytological and
243
expression profiling analysis. As a result, the amount of
starting material available for expression profiling analysis
was extremely limited, so that we initially produced the array
in the nylon membrane format to take advantage of the
superior sensitivity of radioactive labeling and detection.
Instead of omitting individual genes from the analysis to
achieve this purpose, we opted to apply principal component
analysis to the data, resulting in a set of combined features
representing weighted combinations of all genes in the data set.
This approach is far less sensitive to outliers or hybridization
artifacts in individual diagnostic samples, thus increasing the
reliability of the analysis. Robustness of classification was also
the rationale for choosing linear discriminant analysis (LDA)
for the construction of the classifier.
We were able to demonstrate that expression profiling analysis
using our specialized diagnostic array in conjunction with
conventional cytology significantly improves the accuracy of
diagnosis and is especially useful in the classification of
otherwise ‘non-diagnostic’ samples, i.e. samples with low
cellularity or complete absence of intact cells.
Fig. 2.