Diapositiva 1

1
HISTOLOGICAL QUALITY CONTROL OF FROZEN SAMPLES
USING MATCHING FFPE TISSUE PRELIMINARY
RESULTS OF OUR INSTITUTIONAL BIOBANK E.
Mattioli, E. Foglia Manzillo, V. Rubini, F.
Palma, A. Paradiso, G. Simone
2
QUALITY CONTROL IN BIOBANKING

• Definition Quality Control (QC) is a complex of
  procedures aimed at verifying the suitability of
  the collected samples for research
• International publications provide guidelines
  and technical details
• ISBER, 2012 Best Practices for Repositories,
  www.isber.org
• IARC, Technical Standards and Protocols for
  Biological Resource Centres, 2009
• Mager SR et al, Standard operating procedure for
  the collection of fresh frozen tissue samples,
  Eur J Cancer 2007
• Morente MM et al, TuBaFrost 2 Standardising
  tissue collection and quality control procedures
  for a European virtual frozen tissue bank
  network, Eur J Cancer 2006

3
TYPES OF QUALITY CONTROL

• histological
• molecular

2 levels of QC
4
HISTOLOGICAL QC

• AIMS
• verification that the Biobank sample is
  representative of the whole lesion
• verification that the Biobank sample possesses
  adequate morphological and immunohistochemical
  features

MEANS observation of frozen sections of the
banked sample and/or sections from matching
paraffin blocks
5
MOLECULAR QC

• AIM
• verification that the banked samples are
  suitable for research use (preservation of
  relevant molecules)

MEANS assay of the degree of fragmentation of
nucleic acids (DNA/RNA) via extraction,
amplification (PCR) and gel electrophoresis
6
DESIGN OF THE STUDY

• QC PROCEDURE evaluation of critical
  immuno-morphological features on frozen biobank
  aliquots and on matching FFPE material
  representative of the whole lesion
• tumor cellularity
• MIB-1 staining
• AIM testing the concordance (i.e.
  reproducibility) of these features on sections
  from frozen tissue and from the corresponding
  paraffin tissue blocks.

7
1/2
MATERIALS METHODS
60 cases from our Institutional Biobank were
enrolled, including both primary and metastatic
tumors.
Site/type n.
Primary tumors breast 29
Primary tumors colorectal 9
Primary tumors gastric 5
Primary tumors endometrial 5
Metastatic tumors nodal metastases from breast 6
Metastatic tumors liver metastases from colorectal 6
Total 60
8
2/2
MATERIALS METHODS

• At our Pathology Department, as a routine
  procedure, each biobank-dedicated tissue sample
  is cut in two mirror-matching halves
• one half is reduced into aliquots, which are
  weighed and frozen ? _at_Institutional Biobank
• the other half is sent to routine histological
  processing ? specular paraffin block
  (_at_Pathology Unit)

9
1/5
RESULTS
The comparison of frozen aliquots with specular
FFPE tissue showed a variable degree of
discrepancy, due to the intrinsic heterogeneity
of tumor tissue.
10
2/5
RESULTS
The fraction of concordant cases significantly
varied according to the amount of deviation
tolerated in the definition of concordance.
Tolerated deviation Tumor cellularity MIB-1 staining
Low 10 5
Intermediate 15 10
High 20 15
Tumor cellularity MIB-1 staining
Low tolerance 55 (33/60) 52.1 (25/48)
Intermediate tolerance 63.3 (38/60) 68.7 (33/48)
High tolerance 71.7 (43/60) 81.2 (39/48)
11
3/5
RESULTS TUMOR CELLULARITY
Concordance appeared to be influenced by anatomic
site, histotype and cytoarchitectural features
Tumor cellularity - concordance
Tumor group Low tolerance Intermediate tolerance High tolerance Discordant
Endometrial cr 80 100 100 0
Breast cr 51.7 58.6 65.5 34.5
Colorectal cr 44.4 55.5 77.7 22.3
Gastric cr 60 60 80 20
Node metastases (breast) 50 66.7 66.7 33.3
Liver metastases (colon) 66.6 66.6 66.6 33.4

• Endometrial carcinomas and liver metastases
  showed the highest concordance fractions in tumor
  cellularity.

12
4/5
RESULTS MIB-1 STAINING

• Metastatic tumors showed the highest concordance
  fractions in MIB-1 staining

MIB-1 staining - concordance
Tumor group Low tolerance Intermediate tolerance High tolerance Discordant
Endometrial cr 20 40 60 40
Breast cr 55.2 75.9 86.2 13.8
Colorectal cr 40 60 80 20
Gastric cr 50 50 50 50
Node metastases (breast) 66.6 66.6 100 0
Liver metastases (colon) 100 100 100 0

• gastric carcinomas were discordant in 50 of
  cases.

13
5/5
RESULTS
MIB-1 was non-assessable in 20 of cases (12/60)
because of technical artefacts on the sections
from frozen aliquots.

• this finding may indicate a higher lability of
  the molecular components compared to
  morphological attributes.

14
CONCLUSIONS

• immunomorphological features were not as
  reproducible between matching frozen and FFPE
  tissue as expected, because of the inherent
  heterogeneity of tumor tissue.
• mirror-matching FFPE blocks can still play a
  role in biobanking QC procedures
• however, sample collection and storage
  procedures need to be optimized with the intent
  to minimize the effects of tumor heterogeneity
  and to improve the preservation of molecular
  attributes.

15
THANKS!
16
(No Transcript)
17
WAIT!
WHY MATCHING PARAFFIN SECTIONS???

• easier to cut
• better morphological detail/image quality
• no need to thaw samples!

18
(No Transcript)