Dicentric Assay

Post on 02-Feb-2016

49 Views

Category:

Documents

0 Downloads

Preview:

Click to see full reader

DESCRIPTION

Dicentric Assay. Lecture Module 4. Dose response curves (1). 250 kVp X-rays. Neutron mean energies, MeV. 2.0. 0.7. 7.6. 14.7. 1 Gy min -1. 1.5. Dicentrics per cell. 0.5 Gy min -1. 1.0. 0.2 Gy h -1. 0.18 Gy h -1. 0.5. 60 Co g -rays. 0.0. 0. 1. 2. 3. 4. 5. Dose, Gy. - PowerPoint PPT Presentation

Transcript

IAEAInternational Atomic Energy Agency

Dicentric Assay

LectureModule 4

IAEA15772

Dicentricsper cell

Dose, Gy0

2.0

1.5

1.0

0.5

0.03 41 2 5

Neutron meanenergies, MeV

0.7 7.6 14.7 250 kVp X-rays

60Co -rays

1 Gy min-1

0.5 Gy min-1

0.2 Gy h-1

0.18 Gy h-1

Dose response curves (1)

2

IAEA

Dose response curves (2)

Which radiations to use ?

3

IAEA

Dose response curves (3)

• Background control

• What doses to use ?

• How many data points ?

• How many cells to score ?

• How many donors to use?

Each lab should produce its own calibration curves; do not use other’s

4

IAEA

Cell culture (1)

• Need to score first metaphases (M1)

• 48h to complete first cell cycle

• Choose a medium that is not too fast growth

• Use PHA for mitogenic stimulation

• Monitor cell cycling with FPG staining

5

IAEA

Cell culture (2)

Two basic methods:

• Whole blood culture

• Separated lymphocyte culture

6

IAEA

Cell culture (3)

7

IAEA

Cell culture (4)

Fixation

This is a simple procedure

Basic steps:

• Harvest cells from culture

• Swell cells by hypotonic salt solution

• Fix cells in alcohol : acetic acid mixture

• Wash several times in fixative to remove non-metaphase cell debris

8

IAEA

Slide making

Essential points

• Clean, grease-free slides

• Cold, wet slides

• Laboratory ambient temperature and humidity

• Adjust concentration of metaphases

9

IAEA

Pre-treatment

Prior to staining

• Clean up the background cloudiness

• Brief wash in an RNAse A solution

– full protocol is given in the IAEA Manual

10

IAEA

Slide staining

Fluorescence plus Giemsa stain (FPG)

• Essential to ensure M1 scoring

Giemsa stain

• If confirmed that there are no M2 cells present

• If using early Colcemid culture method

11

IAEA

FPG Harlequin staining - an M2 cell

12

IAEA

Slide scoring

Two methods:

• Conventional by-eye

• Automated microscope assistance

13

IAEA

Scoring at the microscope

The old way The new way

14

IAEA

How many cells to score for overdose case ?

• Depends on dicentric frequency

• Depends on the statistical uncertainties needed

• Depends on urgency for result

• Depends on available skilled scorers

• Depends on availability of a metaphase finder

Essential points

15

IAEA

95% confidence limits on 500 or 1000 cells scored

Dose estimate Confidence No of cells scored

(mGy) limits 500 1000

100 Upper 320 245

Lower <0 16

250 Upper 448 380

Lower 111 141

500 Upper 677 627

Lower 333 383

1000 Upper 1178 1127

Lower 830 881

16

IAEA

Data recording (1)

The count of objects must balance

A metaphase with:

A dicentric plus its fragment still = 46

A centric ring plus its fragment = 47

Each excess acentric increases the count above 46 by +1

Tri-, quadri- centrics etc, are converted to dicentric equivalents

17

IAEA

Data recording (2)

Use a standard score sheet

• The sheet should include scorer and case ID

• Full retrieval of the observations on each cell

18

IAEA

Information storage

Securely preserve

• Data sheets

• Slides

• Surplus fixed cells

• Case file notes

19

IAEA

How to present the dose estimate (1)

• Read the dose off the appropriate dose response curve• This is easy

• Calculate the statistical uncertainties on the dose estimate• This is a little more difficult

• Report the result to the doctor or patient• This can be difficult: do they understand

statistics?

20

IAEA

How to present the dose estimate (2)

21

IAEA

How to present the dose estimate (3)

22

IAEA

How to present the dose estimate (4)

Which method to use?

In practice, in most cases the first and simpler

method is sufficient

23

IAEA

How to present the dose estimate (5)

• Does the ‘customer’ – understand what is a confidence limit?

• The biodosimetry laboratory has to be prepared to explain the result and put it into simple comprehensible terms. Not easy!

24

IAEA

How to present the dose estimate (6)

At low doses

• If the lower 95% confidence limit is negative it can be ignored and only the upper limit is of concern

• Consider presenting the result graphically as a probability distribution

25

IAEA

How to present the dose estimate (7)

26

IAEA

How to present the dose estimate (8)

Presenting the dose as an odds ratio

• This also applies to low, possibly zero, dose events

• It needs two sources of information:

A dose estimate from e.g., a badge

A dose estimate from dicentrics

• Which is correct?

27

IAEA

How to present the dose estimate (9)

Odds ratio

• Badge = 66 mSv

• Dicentrics = 1 in 1000 cells (background?)

• Dicentrics = zero dose with UCL = 100 mGy

• Consider 2 possibilities: zero or 66 mSv

• Odds ratio 4.5 : 1.0 in favour of zero dose

• Easy for a patient to understand

28

IAEA

Conclusions

This lecture has covered the dicentric assay from the aspects of:

• Producing dose response curves

• Processing the lymphocytes

• Making and scoring the slides

• Presenting the dose estimates in a way that can be understood by non-scientists

29

top related