Review of CD4 Technologies
Suzanne Crowe, Burnet Institute
On behalf of
Members of the CD4 Working Group,Forum for Collaborative HIV Research
Laboratory monitoring & the 3x5 programCost of ARV Rx in developing countries ispotentially exceeded by the cost of laboratorymonitoring.
Sophisticated assays often established as part ofclinical trial infrastructure, paid for byinternational grants
HIV+ individuals outside trials must find funds topay for their own monitoring
Monitoring HIV treatment is complex– CD4 and VL– Drug toxicity– Adherence– Drug resistance
New CD4 monitoring tests andtechnologies
Flow based assaysGuava EasyCD4 SystemPartec CyFlowPointCAREPanleukogating technology
Manual assaysDynal immune bead-based assay (microscopy)Coulter immune bead-based assay (microscopy)
In the pipelineLabNow microchip technologySemiBio slide test (microscopy)suPAR (soluble urokinase plasminogen activator receptor)CD4 dipstick for remote care
Dual Platform Flow Cytometry
Slide courtesy of Roland Gohde
Single Platform FlowCytometry
Slide courtesy of Roland Gohde
Pan-leukogate or PLG CD4Methodology
Slides courtesy of Angela Vernon and Meryl Foreman, Beckman Coulter
• Identifies CD4+ lymphocytes based ona pan-leucocyte count
• WBC count (cells/ul) X CD4 events from region B CD45 events from region A = Absolute CD4
• The WBC gate is not affected by EDTAchanges that occur with olderspecimens.
• Hematology lymph % is affected byEDTA, count not reliable beyond 24hours.
PLG CD4 Methodology
Gated on WBC
Ungated
Slide courtesy of Angela Vernon and Meryl Foreman, Beckman Coulter
PLG CD4 Count
(r2 = 0.990)
Glencross et al. Clinical Cytometry, 50:2, 2002
Day 1Day 1
Day 5Day 5
Aged Specimen Performance –Limitations of Scatter Gating
CD4 T cellsLymphs
WBC Gate
Forward Scatter cellular structure lost over time, results ininability to define appropriate gates using scatter alone
Slides courtesy of Angela Vernon and Meryl Foreman, Beckman Coulter
PLG: Aged Specimen Performance
Mixed Model ANOVA for trend over time; p=0.8919 CD4 Count Range: 7 – 1579 cells/Μl; Median CD4 count = 371 cells/μL
Slide courtesy of Ank Gowans, Beckman Coulter and CDC Beijing
n = 39 HIV+ volunteers
0
200
400
600
800
1000
1200
1400
1600
1800
0 1 2 3 4 5 6Days
CD
4 A
bsol
ute
Cou
nt
MeanSpecimensLinear (Mean)
Beijing, China
Summary: PLG CD4NewNew flow cytometry-based method
Based on a pan-leukocyte markeruses a 2-color pre-optimized reagentprovides both CD4% and absolute countsextends sample age beyond 24 hrs to up to 5 daysgood correlation with “gold standard” flowcompatible with most flow cytometers
– with 2 color capability & 488 nm laser line<$6 per test
Licensed by Beckman Coulter from NHLS, South AfricaHigh capacity: good for high volume centralized labs
Guava EasyCD4
Slides courtesy of Jeff Harvey, Tina Baumgartner, Leonard Buchner
Guava EasyCD4Measures absolute CD4 (can measure CD8)Sample volume:
10 µL of whole blood (EDTA)Reagents
10 µL of antibody cocktail– Anti-CD3-PE-Cy5– Anti-CD4-PE
180 µL of Lyse-Fix solutionComponents/Software
Dell LapTop computer includedSoftware includes instrument set-up, dataacquisition and analysis
The Guava EasyCD4 System:
EasyCD4 Softwareand Data Handling
Green Diode Laser
36 cm
32 cm
21 cm
CapillaryFlow Cell
Microliter Cell Samples
Waste Fluid
Dell LaptopWindows 2000
Pentium IV
15.9 kilos with PC15.9 kilos with PC
Add 10uL of antibody cocktail to each tubeAdd 10uL EDTA whole blood to each tube,vortex, incubate 15minAdd 180uL of Lyse/Fix solution, incubate15min During sample incubation, turn on powerand allow 10 minute warm-up Run Guava Check QC procedure (5 min) Adjust (or recall) instrument settings Acquire samples; Analyze results
Guava EasyCD4 Protocol
UCSF-GCRC/GIVI-CFAR Core Immunology Laboratory
R2 = 0.9725
EasyCD4 vs MultitestCD4
0
200
400
600
800
1000
1200
1400
1600
1800
2000
0 200 400 600 800 1000 1200 1400 1600 1800
MultiTest TM CD4
Easy
CD
4 TM
Tri
pli
cate
s
R1
R2
R3
Linear (x=y)
Linear (R1)
Linear (R2)
Linear (R3)
R=0.97
North America – California 3 Site Trial
-200
-150
-100
-50
0
50
100
150
200
0 500 1000 1500
Average Guava & BD MultiTest CD4
Res
idua
l Diff
eren
ce Site1Site2Site3Mean +2SD -2SD
Slides courtesy of Jeff Harvey, Tina Baumgartner, Leonard Buchner
N = 75
Guava EasyCD4 at YRG CARE
Summary Easy CD4 (Guava)
Quick assayReagent cost is low ($3/test)Capital costs high ($45,000)Existing QA panels not compatible withthis technologyMeasures percentage CD4 as %CD3+ TcellsStill undergoing evaluation
Partec Cy-Floweg CyFlow Counter1PCyFlow SLGreen 2PCyFlow SL Blue 5P
Slides courtesy of Roland Gohde
Cy-Flow no lyse - no wash CD4 protocol
3-Step Protocol50µl blood from the patient into a sample tubeadd 10µl of CD4-PE and incubate for 10 minutes at RTin the darkadd 850µl of the no lyse dilution buffer
31
2
Slides courtesy of Roland Gohde
0 500 1000 1500
0
500
1000
1500
r = 0.9899
CD
4 co
unt (
FAC
SCou
nt)
CD4 count (CyFlow)
Cameroon: CD4 Counting - CyFlow vs. FACSCount
Douala and Marua, Cameroon
Cameroon: CD4 - CyFlow vs. FACSCount Bland-Altman Plot
0 500 1000 1500-600
-400
-200
0
200
400
600bias 8 CD4/µl
zero biasmean of difference (bias)
mean -2sd
mean +2sd
diffe
renc
e be
twee
n C
yFlo
w a
ndFA
CSC
ount
(CD
4/µl
)
mean of methods CyFlow and FACSCount (CD4/µl)Douala and Marua, Cameroon
Summary Partec Cyflow
Simple and portable flow cytometric assayNo formal comparisons with samplesprovided by external QA programsavailable at this timeReagent costs $2-3/testCapital costs approx $20,000Absolute CD4 only
PointCARE
Slides courtesy of Cecil Sherrer
PointCARE System
%CD4 and Absolute CD4WBC, LY% and countMobile; battery backup
Room temperature reagent storage and operation
4. Lysing Reagent Tube or Cleaning SolutionTube
3. Rinse Tube
2. CD4 Reagent Tube
1. Patient Whole Blood Sample Tube
•Patient sample and reagents bar-code are tracked in the instrument.
Closed- tube operation –biohazard containment via cap piercing
•Ideal for low-volume, decentralized labs
Automated Patient Results
•Both CD4% & AbsoluteCD4
•without beads•critical for pediatrics
•Depending on test volume, cost of patient result is under US$10•Cost of patient result includes:
All reagents and disposablesOperator timeCD4, CD4%, WBC, LY, LY%Service
CD4 Count- Method Comparison N = 68
y = 1.0718x + 30.29
0
400
800
1200
1600
2000
0 500 1000 1500 2000
REF Dual Platform CD4 Counts
Poin
tCA
RE
CD
4 C
ount
s
R = 0.98
CD4 Count- REF 150-350/uL Bias Plot
N = 23
-100
-50
0
50
100
150
150 170 190 210 230 250 270 290 310 330 350
REF Dual Platform CD4 Counts
% B
ias
= (F
C - R
EF) /
REF
PointCARE comparison with DP Flow
Summary PointCARE
Recently received FDA approval%CD4 (as percentage of CD3+T cells)Cost approximately $10/testCapital cost approx $15,000-20,000
Manual low cost assays formonitoring CD4
Vidal, Omah Mooleedhar, Shahir Ali, CAREC Data from Crowe lab, BurnetInstitute Melb and
Dr Balakrishnan’s lab, YRG Care ChennaiArlene Darmanie, Cecile Goddard
CD4 manual methodsDynal assay
CD4Dynabeads®
+ -
Monocyte-depletedblood removed to new tube
+ -+ -
Lysis andstainingof nuclei
Countstainednuclei
CD14Dynabeads®
CD4 cytospheres® Add blood to staining solution Count cells
with beadsattached
Coulter assayMonocyteblocking agent
Crowe, et al. CID 2003: 37 (suppl 1) S25-35
What equipment is needed for thesemanual CD4 assays?
Microscope with 40x objective Hemocytometer 0.1 mm deep Manual counter Tubes Pipettes
Plus rotating wheel andmagnet for Dynal assay
Bland-Altman plot for difference against mean forCD4+ T-lymphocyte counts by flow cytometry and
Coulter cyto-sphere assay(n=122).
Balakrishnan Pachamuthu et al
Mean
120010008006004002000
Difference
300
200
100
0
-100
-200
-300
Dynal assayshows
excellentassociation
with flowcytometry
Correlation of Flow SP and Dynal assay
0
500
1000
1500
0 200 400 600 800 1000 1200Flow cytometry (cells/ul)
Dyn
al a
ssay
(c
ells
/ul)
R= 0.97, n=54
Bland-Altman Plot
-150-100-50
050
100150200250300350
0 200 400 600 800 1000 1200
Average CD4 count by Flow cytometry and Dynal assay (cells/ul)
Diff
eren
ce in
CD
4 co
unt
(cel
ls/u
l)
n=54
Average Flowcytometry resultis 65 cells/µlhigher thanDynal assay
Crowe et al 2003
Comparison of CD4 Count between DYNALand FACSCount
(n=85)
-200
-150
-100
-50
0
50
100
150
200
0 200 400 600 800 1000 1200 1400 1600
Average of Dynal IFA and FACSCount
Res
idua
l (D
ynal
-FA
CSC
ount
)
Series1
Mean
Mean + 2STD
Mean - 2STD
Arlene Darmanie, et al CAREC
Coefficient of variationDynal assay vs flow cytometry in West Africa
13014202449716n
8.49.85.36.09.77.014.6CV
AllSites
Site6
Site5
Site4
Site 3Ref site
Site 2Site 1Sites
DYNABEADSR
Flow Cytometry
4520n
19.78.3CV
Sites Distant fromReference site
Reference site
Diagbouga S et al AIDS 2004
Impact of the delay in sample handling onDynabeadsR Technique at room temperature
samples exhibiting ≥ 20% decrease in CD4 cell counts
50.0 (30.7, 69.4)14 / 28Hour 24
17.9 (6.1, 36.9)5 / 28Hour 12
10.7 (2.3, 28.2)3 / 28Hour 8
0 (0, 12.3)0 / 28Hour 4
% (95% CI)nTime
Diagbouga S et al AIDS 2004
Blood stabilizers
Slides courtesy of Viv Granger and Dave Barnett, NEQASUK and data from Amanda Lindholm Streck Laboratories
Reagents for stabilizing blood samples Guidelines, CD4+ T cell analysis: must be done within 18 hrs
haematology analysers: difficulty producing a differential after 24 hrs
CytoChexTM (Streck laboratories) Member of family of non cross-linking fixatives Orig. designed to preserve WBCs in whole blood (1:1) Cyto-Chex BCT contains 57ul preservative/anticoagulant Samples stable 7 days at ambient temps
NEQAS (UK) TransFixTM lasts >10days, (1:10), <250 C Transfix: allows transportation of fixed samples
Both compatible with flow technology No data on stabilized blood and manual CD4 counts
Turpen & Collins Amer Clin Lab 1996 15:30; Barnett et al Cytometry 1996 26:216Jani et al J Imm Meth 2001 257:145
Flow Flow CytometricCytometric Analysis Analysis
Fresh Day 7
Normal s Unfixed TRANSFix Specimens
Preservation of Lymphocyte Subsetswith TransFIX
0
10
20
30
40
50
60
70
80
Day 1 Day 2 Day 3 Day 4 Day 6
Time after collection
% L
ymph
ocyt
es
CD 3
CD 4
CD 8CD 16/56
CD 19
Stability of CD3/4+ TCells postaddition of TransFIX
Stability of CD3/4+ T Cells post addition of TransFIX
0
200
400
600
800
1000
0 1 10
Days post addition of TransFIX
CD
3/4+
T C
ell c
ount
(c
ells
/uL) <100
>100 but <300>300 but <500>500
CD4
In the pipelineLabNow– Microchip, reader and digital camera– Individual biochips– Absolute CD4– Tentatively planned for availability late 2005
Semi-Bio manual slide technology– Slide with an antiCD4 coated chamber that traps CD4+
cells during incubation– Count CD4+T cells after staining
Remote point-of-care technology– No technical skill required– For estimation of CD4+ T cells– Early stages of development
Microchip technology with a reader device and digital camera
Data/Slide from Bill Rodriguez
Which low-cost CD4 assay to introduce?
Depends onNumber of samples per day– Low throughput, manual may be most cost-effective– High throughput, flow method most cost effective
(and definitely more practical)
Sophistication of lab– Current methods require varying degree of technical
skill– Remote point of care methods under development
Availability of technical support & equipmentmaintenance training– A key issue for flow cytometers– Remote area, opt for manual or ship samples or
ensure local engineers/technicians trained
Cost
Word of caution whenconsidering cost
Hidden costs (ie non-kit costs) greatly influence thefinal cost to a testing laboratory in a resource-constrained country Labour (often less expensive) Disposables (if available often more expensive) Shipping costs Importation costs Infrastructure Repeat assay runs Instrument repair
Pricing may be best negotiated at an international orcountry level with bulk procurement schemes
Elbeik et al Expert Rev PharmacoEconomics Outcomes Res 2003 3:383-407
Where are we up to?All assays/methods continuing to undergo in-country analysesRigorous independent evaluation, includingclinical trial evaluation and independentstatistical analysis, is absolutely essentialSome technologies /assays recently licensed inUSA (PLG, PointCARE)However all are still emerging technologiesQA participation essential and establishing thisshould be part of the dealCountries should not purchase technologiesthat have not been inadequately validated
Final thanks to Forum for Collaborative Research
Ben ChengHoutan MovafaghBen CollinsVeronica MillerAlan Landay (Chair)All those who provided slides for this presentation, especially
• Rolande Gohde (Partec)• Jeff Harvey, Tina Baumgarten, Leonard Buchner (Guava)• Angela Vernon and Meryl Foreman, (Beckman Coulter)• Ank Gowans, Beckman Coulter and CDC Beijing• Dr. Debbie K. Glencross and the NHLS of South Africa• Dr Balakrishnan, YRGCare, Chennai• Douala and Marua, Cameroon• Viv Granger and Dave Barnett, NEQAS UK• Cecil Sherrer (PointCARE)• Vicki Greengrass, Mandy Dunne, Megan Plate, Pauline Steele
(Burnet Institute)• Roche (SC support to attend CROI)