ANALYSIS OF BODY FLUIDSCOMPARISON OF THE XN-BF MODULE WITH

MANUAL METHODS

Kevin Fomiatti

Medical Laboratory Scientist

Wellington SCL – Hutt Laboratory

Topics to be covered

■ Background information

■ XN-BF Module

■ Red cell counts compared

■ White cell counts compared (including eosinophils)

■ WBC differential counts compared

■ XN-HF cells – comparison with cytology results

■ Where to from here?

■ Summary of main points

Background

• Hutt lab - body fluids are processed by core staff who provide a 24/7 service.

• Core staff:

• Cover multiple areas outside of normal work hours

• Processing body fluids can be time consuming

• Not all staff are trained to do differential counts.

• XT-2000i analyser upgrade to the XN 2000. Dedicated body fluid module purchased – envisaged

may streamline the processing of body fluids

• Samples tested August – September 2017

• Result comparison prior to full validation as a preliminary assessment of analyser capabilities –

aid to validation planning

• Both methods treated as equal. Results compared and consider limitations of each method

and causes of potential differences.

Manual Method

• Neubauer chamber for all cell counts

• Cytospin - Diff and Gram based on the table below.

• Red cell and white cell counts reported for all samples, units x 106/L

• Diff counts - % PMN, %MN % Eosinophils.

Samples Tested

• > 17 hour old samples

• mostly tested in Wellington and transferred to Hutt the next day for comparison testing

• Largely omitted from the comparisons due to insufficient opportunity to repeat the manual counts and

due to uncertainty as to how the delay may have affected the diff and red cell counts.

• Samples with cytology also requested were included for comparison with the HF-BF research

parameters.

XN BF Module – Operation & Results

Closed cap – 1 ml

Low RBC Results ComparedAll fluids 1-4 hours old, RBC-BF2 vs Manual Count.

7 points with

manual RBC

counts <10

Red Cell Counts

■CSF Samples:–The RBC-BF parameter lacks the necessary

accuracy and precision to discriminate between low red cell numbers in tubes 1 and 3.

■Other fluids–XN analyser is potentially very useful for obtaining

cell counts and differential counts on bloodstained fluids.

–Red cell counts on fluids other than CSF is of questionable clinical significance

White Cell Count comparison

Manual count initially

12. Repeated using

phase. Refractile and

non-refractile cells

seen.

White Cell Counts

■ Comparison’s at low levels appears promising

■ Variability within the analytical range ( up to 10,000 x 106 cells/L), see possible causes

below.

■ Comparison may be affected by multiple factors, e.g

– manual counts – WBC count rather than a total nucleated cell (TNC) count, therefore

difficulty accurately distinguishing cell types in the chamber, such as mesothelial cells,

histiocytes, macrophages and monocytes. Pipette errors, cell dilution and calculation

errors, type of microscopy (white light vs phase contrast), crenated RBC vs NCs, inter-

observer variation etc.

– XN Calibration – Roche target for XN Check L1 control – 2,950 x 106/L, XN results

consistently above the target and averaging around 3,070 x 106/L, i.e. 4% positive bias.

XN-BF controls were not used and linearity has not been checked by correlating manual

counts against machine counts for a series of dilutions.

WBC Differential – WDF Channel

Whole Blood Body Fluid

Eosinophils

■ XN overestimates on low

counts – probably represents

poor discrimination between

neutrophils and eosinophils for

many of the samples.

■ XN results are lower for

samples where eosinophils

were seen on microscopy

review.

■ Insufficient samples with

eosinophils for proper

assessment. Requires further

assessment with appropriate

cell concentrations on the

cytospin preparations.

XN-BF WBC Differentials – Points to consider

■ XN-BF differential re MN% and PMN% consistently gave comparable results with manual

differentials.

■ Eosinophil count/percent – useful only if there is good discrimination between the

neutrophils and eosinophil regions on the WDF scattergram

Issues for Future Consideration

■ Manual count – WBC or TNCs (Total nucleated cells) “All nucleated cells should be counted,

since it is difficult to accurately distinguish cell types in the chamber” ⊛

■ “Differential counts should include all cell types and be reported as percentages.”⊛

■ “The term mononuclear cell should be avoided, since the term does not adequately

distinguish monocytes from lymphocytes, a distinction that has diagnostic significance.” ⊛

⊛ (Clinical and Laboratory Standards Institute. Body Fluid Analysis for Cellular Composition; Proposed Guideline. CLSI document H56-P

[ISBN 1-56238-575-5]. Clinical and Laboratory Standards Institute, 940 West Valley Road, Suite 1400, Wayne, Pennsylvania 19087-1898

USA, 2005.)

HF Cell Counts

■ HF Cells contain large

amounts of nucleic

acids and so form a

clearly discernible cell

population toward the

top of the

scattergram (high

fluorescence).

■ Cell types include

macrophages,

mesothelial cells and

malignant cells.

High Fluorescent (HF) Cell % Comparison with Cytology

Results• Cytology Report – assigned

categories

• Unremarkable - Normal or

predominantly WBC.

• Reactive – mesothelials or reactive

mesothelials present, no malignant

cells seen

• Malignant – malignant cells, with or

without reactive mesothelials.

• HF cells > 7.5 % - predominantly

reactive or malignant.

• Cannot distinguish reactive from

malignant samples based on

HF%

• Ascitic fluid from an 82 yo

patient presenting to ED on

20/10/17

• No clinical details provided, no

specific tests requested.

• Sample registered for

microscopy and culture.

• XN-BF results – RBC 500, WBC

197, 3% PMN, 97% MN, 39% HF

• Reported WBC 172, RBC 255,

1% PMN, 99% MN, No growth.

• Slide reviewed by a pathologist,

cytology added on.

Case Study

Malignant cells from a

metastatic adenocarcinoma.

Where to from here? – proposal

• Purchase the CLSI Standard and conduct a gap analysis between what we do and

what the Standard recommends.

• Process all body fluids via optimised manual methods and on the XN analyser to

build up comparative data and to investigate the causes of any discrepancies that are

identified. Rather than creating additional work for staff, utilise the XN analyser to

streamline the manual processes as per the next slides.

• Continuing education sessions and training for staff

• Concurrently verify Roche’s claimed performance characteristics for the XN-BF

module. e.g precision, accuracy, linearity, carryover, analytical range, sensitivity etc.

• Report on findings and recommend changes to improve the procedures

Spreadsheet for XN-Results

• Enter WBC and RBC from XN Analyser

• Adjust Sample and Diluent volumes to achieve optimal numbers in a Neubauer chamber

• Print off the form shown on the next slide

• Previous spreadsheet data populates the fields

• Micro lab worker

• sets up dilutions as provided for WBC and

RBC

• Sets up cytospin preps as per the table

• Record the cell counts in the above table

Summary of Main Points

XN-BF Module• Differential re PMN% and MN% - excellent correlation with manual differentials

• Low red cell counts – unreliable as a means of establishing whether low numbers of

RBC in CSF are a result of a traumatic tap.

• Eosinophils – research parameter which should not be used when there is poor

discrimination between the neutrophil and eosinophil clouds on the WDF scattergram.

• HF-BF% - research parameter. Results suggest that percentages above 7%, especially

on serous fluids, may indicate either a reactive process or malignancy.

• Cytospin – XN cell counts have proved useful in calculating a specific volume of fluid to

use for the cytospin preparation in order to optimise cell numbers across the range of

cell counts.

• White cell and red cell counts – further work required.