1994/1995 COMMITTEE MEMBERS

Diana Benn Jeanette Drew Karen Holdaway

Peter Hobson Christine Smyth

1996 COMMITTEE MEMBERS

Mr Peter Dynes Mr Peter Hobson Dr Ian Taylor

9. DNA INVESTIGATIONS

CONTENTS

9.1 Introduction

9.2 Specimen

9.3 Specimen Collection, Transport and Integrity

9.4 Specimen Processing

9.5 Controls

9.6 Flow Cytometer Quality Control

9.7 Sample Analysis

9.8 DNA Histogram Interpretation

9.9 Data Reporting

9.10 References

9.11 Appendix: Examples of DNA Histograms from Frozen Malignant Solid Tumours

9.1 INTRODUCTION

DNA analysis by flow cytometry is a rapidly expanding technology that is moving from the research laboratory into the clinical laboratory. Recent advances in the promotion, availability and increased usage of this technology, have clearly created a need for procedural guidelines and proficiency testing programs.

International attempts have been made to identify the problems associated with quality assurance for DNA analysis by flow cytometry. This resulted in the production of DNA consensus documents^1­7 which serve as guidelines for laboratories wishing to implement flow cytometric DNA ploidy and cell cycle analysis studies.

This document has been designed to aid scientists in the implementation of comparable practices between laboratories^5,8, thereby leading to the establishment of national quality assurance procedures. The DNA Subcommittee recognises that as there are many types of flow cytometers. This document addresses only those issues common to all. Therefore, at this stage, all encompassing guidelines are beyond the scope of this document.

9.2 SPECIMEN

All specimens for DNA content analysis by flow cytometry need to be prepared either as single cell suspensions or as isolated nuclei before running on a flow cytometer. DNA analysis may be performed on peripheral blood, bone marrow aspirates, tissues (fresh or frozen), archival biopsies (formalin­fixed/paraffin­embedded) and cytological specimens (e.g. urine, aspirates, cervical smears)^9-11.

9.2.1 Fresh and unfixed material for cellular DNA analysis is preferable to formalin­fixed/paraffin­embedded material. In most cases high quality single cell suspensions can be obtained from fresh tissues (solid or fluid), which are considered the optimal tissue sample¹².

9.2.2 Fresh/frozen tissue often proves to be more practical for routine clinical flow cytometric DNA analysis because of the difficulty of ensuring adequate handling and storage of fresh specimens from the time of surgical excision or collection. Consideration should be given in deciding whether to freeze all biopsies before analysis as a matter of course because fresh tissue is often required for multi-parametric analysis. Multi­parametric DNA analysis has not been addressed in this document.

9.2.3 Paraffin­embedded biopsies yield only bare nuclei from the disaggregation treatment to produce single nuclear suspensions for flow cytometric DNA analysis^13,14. The histograms produced tend to have broader peaks (higher coefficients of variation ­ CVs) and increased cell debris when compared to fresh tissue preparations¹⁵. Using paraffin­embedded tissue has some advantages over the analysis of fresh tissues. Obtaining a separate piece of fresh tissue for analysis by flow cytometry is often difficult and may be impossible when:

• The biopsy of a solid tumour is very small.
• The need for flow cytometric analysis is not apparent until after all of the tissue has been embedded or discarded by the pathology laboratory.
• The tumour is difficult to identify in fresh tissue, e.g. prostatic chips. Examination of H&E sections will pinpoint the areas of the highest tumour density.
• A retrospective study of patients whose clinical course is already known needs to be performed. Prospective studies on fresh tumours could take years to accumulate data to determine correlation statistics.

9.3 SPECIMEN COLLECTION, STORAGE, TRANSPORT AND INTEGRITY

The nature of the collection and transport of the specimen will vary depending on the type of specimen. Recommendations for the major specimen types are listed below:

9.3.1 Sample Collection Conditions

9.3.1.1 Blood and bone marrow aspirates can be anticoagulated with ACD, EDTA or heparin. Universal precautions should be strictly observed when collecting blood samples (See 1.1 Safety Guidelines). Specimens may be kept at room temperature for up to 24 hours and thereafter at 4°C. Analysis of specimens over 72 hours old is not recommended. If the samples are to be frozen for delayed analysis, it is necessary to remove erythrocytes before freezing (e.g. by gradient centrifugation or hypotonic lysis).

9.3.1.2 Cervical smears should be dispersed into cold, physiological medium for subsequent investigation.

9.3.1.3 Fine needle aspirates should be collected into cold physiological medium on ice to reduce deterioration of the specimen.

9.3.1.4 Fresh tissues should be collected into a clean container with abundant cold physiological medium kept on ice. Fresh tissue may be frozen in native state immediately after excision, or in tissue culture medium depending on tissue type and size.

9.3.1.5 Paraffin­embedded (archival) tissues: The effect of many different fixatives and preparation protocols have been widely examined and must be recognised when analysing and interpreting results from such samples. Neutral buffered 10% formalin is the fixative of choice for flow cytometric (FCM) DNA content on paraffin­embedded tissues. Fixation with Bouin's or Zenker fixatives results in poor to uninterpretable DNA histograms.

9.3.1.6 Pleural aspirates/lavage fluids may require anticoagulation. Store on ice.

9.3.1.7 Urine/bladder washings should be stored on ice prior to analysis, and process within 24 hours.

9.3.2 Sample Storage Conditions

If fresh specimens cannot be analysed within 24 hours of disruption from the host, process the specimens for freezing. The recommended freezing options for fresh tissue and cell suspensions are: ­20°C: for overnight or short term storage (<4 weeks)

­70°C or liquid N₂: for short to long term storage (>1 week)

Freezing medium (e.g. DMSO) is excellent for preserving cell suspensions¹⁶.

9.3.3 Sample Transport Conditions

Packaging, labelling and transport of specimens should comply with all current local, state, national and international regulations for the regions through which the specimens will pass.

9.3.3.1 Blood and bone marrow aspirates: Transport at room temperature.

9.3.3.2 Other fresh specimens: Transport chilled on ice.

9.3.3.3 Frozen specimens: Transport on dry ice or in liquid nitrogen.

9.3.3.4 Paraffin­embedded (archival) tissues: Transport at room temperature.

9.3.4 Specimen Integrity

It should be ensured that specimens comply with appropriate collection, transport, storage and integrity requirements for flow cytometric DNA analysis before proceeding with processing.

9.3.4.1 Blood and bone marrow samples should not be haemolysed or clotted. Analysis of specimens greater than 3 days from time of collection is not recommended.

9.3.4.2 Cervical smears should be examined for clumping, degeneration or autolysis by microscopy. Analysis of specimens greater than 24 hours from time of collection is not recommended.

9.3.4.3 Fine needle aspirates should be examined for clumping, degeneration or autolysis by microscopy. Analysis of specimens greater than 24 hours from time of collection is not recommended. Fresh cells from needle aspiration biopsies have to be rapidly processed to minimise cell clumping, cell deterioration and to maximise tumour cell yields.

9.3.4.4 Fresh/frozen tissues should be examined macroscopically for suitability for DNA analysis by ascertaining the presence of sufficient representative areas of the tissue/tumour. Avoid necrotic, fibrotic or fatty tissue components. If no other more suitable area for sampling exists, care should be taken in interpreting histograms derived from such specimens.

9.3.4.5 Paraffin­embedded tissues: Examine an H&E section of the biopsy microscopically for its suitability for analysis. Avoid areas of necrosis, fibrosis or inflammation. Focal areas of interest (tumour) may be selectively removed for analysis by scoring with a scalpel, taking punch biopsies of the blocks, or by careful separation of these areas after the sections have been cut.

9.3.4.6 Pleural aspirates/lavage fluids should be examined for clumping, degeneration or autolysis by microscopy. Analysis of specimens greater than 24 hours from time of collection is not recommended due to the high protein content of these fluids.

9.3.4.7 Urine/bladder washings should be examined for degeneration or autolysis by microscopy. Analysis of specimens greater than 24 hours from time of collection is not recommended.

9.4 SPECIMEN PROCESSING

9.4.1 Irrespective of the source, the tissue samples used for flow cytometry must be shown to contain adequate neoplastic material before clinical samples are processed for cytometry. At the DNA Cytometry Consensus Conference held in 1992, it was recommended that tissue samples should contain a minimum of 20% tumour cells and higher proportions are advisable if tumour proliferation calculations are needed⁵.

For body fluid samples and washings, less than 20% tumour cells may be present and still be adequate for DNA ploidy analysis.

9.4.2 Enrichment for tumour or cells of interest may be necessary by dissection of regions of representative tumour from selected areas of the tissue sample (fresh or paraffin­embedded), or by purification methods for blood or bone marrow specimens.

9.4.3 Multiple samples for DNA ploidy may be necessary in some tumours due to tumour heterogeneity.

9.4.4 Retain a permanent morphologic record of the tissue used for the isolation of cells or nuclei for flow cytometry in the laboratory (e.g. blood film, touch imprint, H&E section, cytospin).

9.4.5 Specific methods for sample handling and preparation for analysis and storage will vary depending on the specimen type, tumour system and methodology. It would be presumptive for these guidelines to suggest procedures which would encompass all specimens. It is the view of the DNA subcommittee that laboratories performing these analyses communicate with others performing similar studies and keep abreast of literature in their areas of interest^17-20.

9.5 CONTROLS

9.5.1 Fresh/Frozen Tissue

DNA diploid reference cells should always be used to identify the position of the DNA diploid G0/G1 peak on the DNA histogram. "The ideal reference cells are diploid cells from the same tissue and the same individual"¹, in that both the chromatin structure and the DNA stainability most closely parallel the cells of interest. This is not always practical and the following suggestion provides an alternative.

Purified preparations of normal peripheral blood mononuclear cells (PBMCs), frozen in small aliquots, are an acceptable and practical standard for DNA investigations.

For reliable quantitation of DNA, identical processing is essential for the tumour and the tissue control. The reference cells should be mixed with the sample before staining when used as an internal standard. The sample should also be run without reference cells.

Chicken red blood cells or trout red blood cells may be used for instrument calibration but are not appropriate for calculating the DNA index.

9.5.2 Paraffin­Embedded Tissue

Paraffin­embedded, normal human tissue (e.g. lymph node, spleen) should be used for inter­run performance assessment of staining.

For each specimen, adjacent non­malignant tissue within the paraffin block is the most suitable DNA diploid reference standard. Where this is not available, non­malignant tissue from another patient may be used. It must have been fixed and paraffin­embedded in an identical manner at the same time as the tumour specimen of interest. Mixing of the DNA diploid standard and the tumour sample should not be performed⁵.

Freshly prepared PBMCs or non­mammalian nucleated erythrocytes are not suitable as diploid reference standards.

9.6.0 FLOW CYTOMETER QUALITY CONTROL

Refer to section 3 Flow Cytometer Quality Control in particular section 3.3 Linearity.

9.7 SAMPLE ANALYSIS

9.7.1 Data Acquisition

Instrument settings during the sample data acquisition (e.g. sample flow rate, voltage etc.) should not be changed.

It is recommended that the data acquisition rate is less than 200 events per second.

9.7.2 Number of Cells

To date, there has been no agreed upon criterion for the number of cells required for generating an adequate DNA histogram. As a rule, 20,000 cells represent the desired quantity, however, ploidy information and in some cases, accurate cell cycle data may be obtained with fewer cells. 5,000 cells are considered the absolute minimum for any interpretation. The object of acquiring larger numbers of cells is to reduce statistical fluctuations in the histogram.

9.7.3 Range of Data

As much information as possible should be gathered on the sample whilst data is being acquired by the cytometer. This may include time as a parameter. Time versus fluorescence gives valuable information about sample flow rates and instrument performance. Exclusion of data seemingly erroneous at the time may be required at a later date. Data above the G2 of the population with highest ploidy may contain valuable information relating to the degree of aggregation and DNA aneuploid hyperdiploid peaks may not be detected if these 'high' channels are discarded or are accumulated in the last 'overflow' channel. As a general rule, observe channels at least 50 percent above the highest G2 peak.

9.7.4 Debris

Debris should be included in the events for analysis ­ not ignored. Do not set a higher discriminator or gate to exclude the debris. A high discriminator will result in falsely elevated S­phase estimates and will affect the accuracy of cell cycle analysis. The debris which is observed in channels below the diploid G1 peak is only one feature of this problem. Other debris will be present in channels which underlie S and G2M phase populations.

There is commercially available DNA modelling software containing several sophisticated approaches for assessing the effects of debris on the cell cycle^21,22. Where debris accounts for >20% of total cells analysed, DNA modelling software should be used.

9.7.5 Doublet Discrimination

The presence of doublets will affect cell cycle analysis and must be excluded by appropriate gating. Most instrument manufacturers provide specific protocols for excluding/minimising doublet contamination. However, this minimising may result in the loss of valuable information.

9.8 DNA HISTOGRAM INTERPRETATION

9.8.1 Definitions

In 1984, the Committee on Nomenclature of the Society of Analytical Cytology published guidelines for a Convention on Nomenclature for DNA Cytometry¹. The recommended definitions were reinforced in the 1993 Guidelines for Implementation of Clinical DNA Cytometry⁵ and are encouraged by this subcommittee for adoption as standard nomenclature.

The terms "DNA diploid" and "DNA aneuploid" should be used, rather than the cytogenetic terminology (hypodiploid, etc.), as no direct measurement of changes in the number or composition of individual chromosomes has been made. The degree of DNA content abnormality is given as the DNA index. By definition a DNA diploid specimen has a DNA index of 1.0.

9.8.2 DNA Diploid

Only one G0/G1 peak is observed. A broad peak described by a large coefficient of variation may obscure a second peak. The coefficient of variation of the G0/G1 peak must be less than 5% for single cell suspensions prepared from fresh/frozen tissues, and less than 8% for nuclear suspensions prepared from fixed, paraffin­embedded specimens. Where a diploid peak only is observed, one should ensure that tumour cells are present in the clinical sample analysed.

9.8.3 DNA Aneuploid

DNA Aneuploidy is reported when at least two separate G0/G1 peaks are demonstrated. For some samples, the diploid/normal peak might be almost nonexistent; hence care should be taken to assign peaks (see 9.11 - Histograms 3a and 3b). Descriptive aneuploid terms may be used for further clarification, but not in replacement of "DNA Aneuploid", ie:

"DNA Aneuploid (Hyperdiploid)" Presence of abnormal peak above diploid peak

"DNA Aneuploid (Hypodiploid)" Presence of abnormal peak below diploid peak

"DNA Aneuploid (Multiploid)" Presence of two or more abnormal G0/G1 peaks

"DNA Aneuploid (Tetraploid)" Presence of abnormal peak at the diploid G2/M position

Histograms are described as DNA Tetraploid when the G2/M fraction exceeds 15%, or at a value determined to be appropriate for a particular organ system. The presence or absence of the corresponding aneuploid G2/M population in the 8N position may be noted. If 6N peaks are noted without a major corresponding 3N peak, this may be an indicator of dumping ie. 6N due to triplets.

9.8.4 Cell Cycle Analysis

The presence of excessive debris, clumping of nuclei, multiploid distributions as well as broad peaks (described by large coefficients of variation) can lead to inaccurate S­phase measurements. There are a number of mathematical modelling programs available to analyse cell cycle compartments^21,22

See 9.11 Appendix for examples of DNA histograms from solid tumours run on different flow cytometers.

9.9 DATA REPORTING

9.9.1 Reports must be produced for all specimens even if ploidy results were unable to be obtained. In this instance, the report must give the reason for the non­assessment.

9.9.2 In addition to the standard components of the interpretative report usually issued by the laboratory, include the following information:

9.3.3 Specimen information ie. site, collection date, analysis date, type of preservative (if applicable).

9.3.4 If several samples have been analysed for investigation of ploidy heterogeneity, the results of each analysis should be reported with a description of the details of sampling.

9.3.5 Proportion of the total population for each sub­population of cells in the cell cycle.

9.3.6 DNA Index

9.3.7 Histogram(s) obtained from the specimen

9.3.8 Interpretation of the histogram using the terminology recommended in 9.8 DNA Histogram Interpretation.

9.10 REFERENCES

1. Hiddemann W, Schumann J, Andreeff M et al. Convention on nomenclature for DNA cytometry. Cytometry 1984;5:445­446.

2. Bauer KD, Bagwell B, Giaretti W et al. Consensus review of the clinical utility of DNA cytometry in colorectal cancer. Cytometry 1993;14:486­491.

3. Duque RE, Andreeff M, Braylan RC et al. Consensus review of the clinical utility of DNA cytometry in neoplastic hematopathology. Cytometry 1993;14:492­496.

4. Hedley DW, Clark GM, Cornelisse CJ et al. Consensus review of the clinical utility of DNA cytometry in carcinoma of the breast. Cytometry 1993;14:482­485.

5. Shankey TV, Rabinovitch PS, Bagwell B et al. Guidelines for implementation of clinical DNA cytometry. Cytometry 1993;14:472­477.

6. Shankey TV, Kallioniemi O, Koslowski J et al. Consensus review of the clinical utility of DNA cytometry in prostate cancer. Cytometry 1993;14:497­500.

7. Wheeless LL, Badalament RA, de Vere White RW, et al. Consensus review of the clinical utility of DNA cytometry in bladder cancer. Cytometry 1993;14:478­481.

8. Bauer KD. Quality control issues in DNA content flow cytometry. Annals New York Academy of Sciences 1993;677:59­77.

9. Coon JS & Weinstein RS (Eds). Diagnostic flow cytometry. Techniques in diagnostic pathology. Academy of Pathology, USA, 1991.

10. Vielh P. Flow cytometry guide to clinical aspiration biopsy. Igaku­Shoin Ltd, USA, 1991.

11. Pallavicini MG, Taylor IW, Vindelov LL. Preparation of cell/nuclei suspensions from solid tumours for flow cytometry, in Melamed MR, Lindmo T, Mendelsohn ML eds. Flow Cytometry and Sorting. New York: Wiley-Liss, 1990:187-194.

12. Bauer KD, Duque RE, Shankey TV (Eds). Clinical flow cytometry: Principles and applications. Williams and Wilkins, USA, 1993.

13. Darzynkiewicz Z, Robinson JP, Crissman HA. Methods in cell biology : Flow cytometry: (2nd ed.) Academic Press Inc. USA, 1994;41(A).

14. Overton WR & McCoy JP. Reversing the effect of formalin on the binding of propidium iodide to DNA. Cytometry 1994;16(4):351­356.

15. Wersto RP, Liblit RL , Koss LG. Flow cytometric DNA analysis of human solid tumours: A review of the interpretation of DNA histograms. Human Pathology 1991:22(11):1085­1098.

16. Foucar K, Chen I, Crago S. Organisation and operation of a flow cytometric immunophenotyping laboratory. Seminars in Diagnostic Pathology 1989:6:13-36.

17. Riley RS, Mahin EJ, Ross W. Clinical applications of flow cytometry. Igaku­Shoin, 1993, USA.

18. Robinson JP (Ed). Handbook of flow cytometry methods. Wiley­Liss Inc, USA, 1993.

19. Shapiro HM. Practical flow cytometry (3rd edition). Alan R. Liss Inc, USA, 1995.

20. Givan AL. Flow cytometry: First principles. Wiley­Liss Inc, USA,1992.

21. Rabinovitch PS. Multicycle enhanced DNA content and cell cycle analysis. University of Washington Phoenix Flow Systems, USA, 1993.

22. Verity: Mod Fit. Operations manual. Verity Software House Inc, USA, 1988-1991.

The following histograms (1 ­ 4) are paired. On the left hand side are the histograms from the tumour single cell suspension and on the right hand side the histograms derived from the tumour single cell suspension to which standard DNA diploid cells (PBMCs) have been added.

1(a) DNA Diploid ­ Breast

1(b) DNA Diploid ­ Breast with added PBMCs.

2(a) DNA Aneuploid (Hyperdiploid) ­ Breast

2(b) DNA Aneuploid (Hyperdiploid) ­ Breast with added PBMCs

3(a) DNA Aneuploid (Tetraploid) ­ Medulloblastoma*

3(b) DNA Aneuploid (Tetraploid) ­ Medulloblastoma with added PBMCs.

* This medulloblastoma specimen was a homogeneous tumour sample, with very few DNA diploid cells.

4(a) DNA Aneuploid (Multiploid) ­ Posterior fossa. The two aberrant sub­populations are calculated separately to produce the two different DIs.

4(b) DNA Aneuploid (Multiploid) ­ Posterior fossa with added PBMCs.

9.11 APPENDIX (Continued)

9.11 APPENDIX (Continued)

9.11 APPENDIX (Continued)

9.11 APPENDIX (Continued)