Upload
others
View
1
Download
0
Embed Size (px)
Citation preview
Point-of Care Technology: Perspective from South
Africa
Prof W. StevensWits University & NHLS
Content
� General Considerations
� South African perspective
� Three feasibility studies• CD4 PIMA
– Validation steps– Sampling issues– Impact on clinical outcome
• GeneXpert TB• Future Viral load
Definition
� Point of Care testing (POCT): refers to testing that is performed near or at the site of the patient with the result leading to a possible change in the care of the patient.
� Outpatient clinic, ER, theatre, mobile clinics, PHC clinics, or even small laboratories
� Small bench top analysers such as blood gas machines or full blood count analyzers, Portable hand held devices such as glucometers, strip based assays
(Warsinke, 2009; Plebani,20009)
Trends in POC
• Fastest growing segment in diagnostics : 10% annually
• 31000 different instruments, 25% of total lab revenue
• Growth areas : cardiac, coagulation, HIV testing , drugs of abuse, tight glycaemic control, molecular diagnostics for infections
• Expanded test menus
• Consolidated platforms e.g. glucose, lactate, creatinine
• Self-contained cartridges
• Improved analytical capability
• Improved data management systems, multi-vendor enterprise solutions
4
Common themes in Research and Development facilitate new assays …
� Microfluidics for fluid delivery
� Minituarization/ low-cost microfabrication
� Magnetism/magnetoresistance
� Efficient, inexpensive light sources (semi-conductor based lighting)
� Affordable microelectronics and digital imaging hardware
� Simplified power sources
� Electromagnetic Actuation of Fluids using Micro/NanoParticles
5
Desirable Specifications of a POC test
� Rapid, results within time-frame of a consultation/clinic visit
� Require minimal training, easy to perform� Simple equipment: Hand held detection
device/strip� Whole blood� Battery operated: an advantage� No specialised laboratory set-up� Stable, temp independent, no cold chain� Affordable, commercially available� Quality control� Ability to transfer data
Advantages Disadvantages� Quality and efficiency of care
can be improved in certain scenarios
� Improved accessibility
� Improve patient compliance and LTFU
� Improved turnaround time
� Smaller sample volumes
� Economic benefits –
� reduced length of stay
� reduced complications and readmission
� Improved patient and clinician satisfaction
� Difficulties with quality control/documentation
� Higher unit of cost/reagent
� Greater personnel requirements at clinic
� Longer patient wait times
� Poor regulatory control
� Data management/audit issues
� Slower sequential processing time/throughput in high clinics
� Over-servicing
7
Current Activities in South Africa
• Component of new National laboratory policy (Maputo declaration)
• Establishment of an NHLS POC working group
• Proposed role in investigation of:– Technologies, equipment– Method validation– Development of SOP’s, training programs and manuals,
checklist– Supportive internal QC and EQA programs– Ethics clearance for clinical studies and pilot sites– CDC support
Recent analysis of rapid HIV VCT testing in South Africa: concerns over performance
Pilot multi-disciplinary POC HIV clinic site
Immediate CD4 count result and retention in care
� Hypothesis : Patients who receive and immediate CD4 count result will more likely present to a facility for ART i nitiationor for wellness.
� Design� Prospective randomized study
� Patients receiving standard of care + immediate CD4� Patients receiving standard of care + information leaflet� Patients receiving standard of care only
� Appointment in one week � No take-home message
� All patients get questionnaire on loss to follow-up reasons � Patients were recruited before their HIV test:
� negative patients are asked about VCT experience (Evaluation of post-test counseling)
� Tests were performed on a BD FACS count machine by lay personnel after safety and technical training
� (R. Osih, 2010)
Patient Characteristics
All Standard of Care
+ Leaflet Immediate CD4 test
Total number 344 119 116 119
Female (%) 223 (64.8) 76 (63.8) 69 (59.5) 78 (65.5)
Age (Range) 31.8 (19-59) 31.1 (19-56) 31.7 (20-50) 32.5 (20-59)
Mean CD4 count (cells/mm3) 346.4 335.4 336.9 364.7
> 215 (%) 232 (67.4) 78 (67.2) 74 (63.7) 80 (67.8)
< 215 (%) 112(32.6) 28 (32.8) 35 (36.3) 39 (32.2)
There were no significant differences in age, gender, employment status and education between the groups
Results
� 107/225 (47.4%) of patients who did not receive an immediate CD4count collected their results after a week
� Patients with CD4 counts < 215 were more likely to report for care if they got an immediate result when testing
Standard
of care
+Leaflet Immediate CD4
count
Reported for care 10 13 23
Did not report for care 18 22 16
Total 28 35 39
35.7% 37.1% 58.9%
P value=0.012
CD4 testing:
� C
13
Beckman Coulter XL
PIMA
Centralised flow cytometry
PIMA CD4
Overview of POC CD4 Technologies1. Rapid CD4 tests = laminar flow strips
BC ZyomyxTM First SIGN™
2. Low volume Fluorescent Cartridge Analyzers
PIMA™ LabNow™ Daktari
3. Low volume Flow Analyzers
FACSCount™ PointCare™
•Screening for ART only•Positive/Negative answer•Cut-off 350cells/µl
•Imperial Cancer funded•Largely prototype•No peer reviewed validation/ publication
•Absolute CD4 count (no CD4%)•10 samples/instrument/day•Biohazard waste-disposal (can be disposed with sharps
waste)•No cold-chain•No peer reviewed validation/ publication•Battery operated but needs mains for charging
•Absolute CD4 count and CD4%, •20-40 samples/instrument/day•Cold chain required for some reagents•Needs special biohazardous waste management•FACSCount is “open” system whereas PointCare is a
‘closed-tube’ system (less biohazardous risk)•“Mini”-laboratory required (only for FACSCount)•Peer reviewed validation/ publication FACSCount
onlyFlow Cytometry/ National CD4 Reference Laboratory, Department of Molecular Medicine and Haematology, CMJAH NHLS
Low volume Fluorescent Cartridge Analyzer systemsPIMATM
• 20min/day internal QC procedure
• Sample analyses 20 minutes
• 10 sample/instrument/8 hour day
• Open blood system/ some risk to operator
PIMATM - PHASE I Laboratory validation:
Fixed volume EDTA venous blood
•Bland-Altman analyses revealed a bias of -16±36 (95%LOA -88 to 55.2)
•%Similarity indicated tight reproducibility agreement of 98%±4.5precision to laboratory based testing(CV) of
4.6% Flow Cytometry/ National CD4 Reference Laboratory, Department ofMolecular Medicine and Haematology, CMJAH NHLS
• Bland-Altman analyses revealed a bias of -38±138 (-16±36 )95%LOA: -388 to 311 , (-88 to 55.2)
• %Similarity showed poorer reproducibility agreement of 99% ±22.7 (98%±4.5) with precision to laboratory based testing
(CV) of 23.0% (4.6%)
PIMATM - PHASE II Well resourced clinic (CMJH)
Finger prick - Capillary bloodN= 95
PIMATM - PHASE III Rural Clinic evaluation: Polokwane
Finger prick - Capillary blood
Phase III clinic data
0 20 40 60 800
200
400
600
800
1000
MeanStd. Deviation
%Sim135.3100.5
Coefficient of variation 74.28%
Sample
%S
imila
rity
Phase III clinic data(1 outlier removed= possible typing error)
0 20 40 60 800
100
200
300
400
MeanStd. Deviation
%Sim123.845.53
Coefficient of variation 36.78%
Sample
%S
imila
rity
Bland-Altman(n=68)
0
500
1000
1500
2000
-800
-600
-400
-200
0
200
400
600
800
Bias
SD of bias
95% Limits of Agreement
From
To
116.4
187.0
-250.2
482.9
Average
Diff
eren
ce
Linear regression
0 500 1000 1500 20000
500
1000
1500
2000 Best-fit values
Slope 0.9610 ± 0.07974Goodness of Fit
R square 0.7111
NHLS
PIM
A
Flow Cytometry/ National CD4 Reference Laboratory, Department of Molecular Medicine and Haematology, CMJAH NHLS
• Bland-Altman analyses revealed
bias of 116.4±187 (-38±138) (-16±36 )
95%LOA: -250 to 483 (-388 to 311) (-88 to 55.2)
• %Similarity showed poorer reproducibility / agreement
123.8% ±45.5 (99 % ±22.7) (98% ±4.5)
with precision to laboratory based testing
(CV) of 36.8% (23.0%) (4.6%)
N=61
PIMA™ Summary ARY
Excellent potential but finger prick sampling introduces error/ CV 5% ↑ >30%
Waste disposal of closed cartridges into biohazard ‘sharps” containers
Instruments able to analyze stabilized blood for external quality control programmes
Internal quality control material available and results indicate very good reproducibility of the system with CV’s <2%
Not for paediatric samples as this require a CD4% of lymphocytes
Nucleic Acid Testing20
GeneXpert for TB
Centralized automated testing500ul processed, 1ml raw sputum
60 days
Evaluation of TB diagnostics in TB suspects compared to culture and clinical outcome
� Sputum from visit 2 � HIV+ (69.6%)
Laboratory comparison to culture (single sputum)
Total
n=286
Smear
microscopy
n=282
MDRPlus
(Hain)
n=251
LightCycler
MTB
n=259
Gene Xpert
n=187
Sensitivity 54.4% (45.2;63.5) 72.1% (62;80) 70.9%(61;79) 84.8%(75;92)
Specificity 99.4%(96;99.9) 97.3%(93;99) 96.6%(92;99) 97%(92;99)
PPV* 98.5%(91;99.9) 94.9%(87;98) 94%(86;98) 95.7%(87;99)
NPV* 74.4%(68;80) 83%(77;88) 82%(75;87) 89.7%(82;94)
Performance in HIV positive individuals
HIV +
n=200
Smear
microscopy
n=196
MDRPlus
(Hain)
n=182
LightCycler
MTB
n=189
Gene Xpert
n=133
Sensitivity 48.7%(37;60) 65.2%(53;76) 63.5%(51;74) 83% (70;91)
Specificity 99.2%(95;99.9) 97%(92;99) 95.6%(90;98) 96%(89;99)
PPV 97.4%(85;99) 93.7%(82;98) 90.4%(78;96) 93.6%(82;98)
NPV 74.5%(66;81) 82%(74;88) 80%(72;86) 89.5%(81;95)
Continued..
Smear + Culture +MDRPlus (Hain) n=42/45 (93.3%) indeterminate
LightCycler MTB (Roche) n=46/46 (100%)
Gene Xpert (Cepheid) n=34/34 (100%)
Smear - Culture +MDRPlus (Hain) n=10/29 (34.4%) 2
indeterminate
LightCycler MTB (Roche) n=12/44 (27.3%)
Gene Xpert (Cepheid) n=17/27 (62.9%)
LightCycler detected n=2 M.avium
72.5% in Boehme, NEJM
Comparison to clinical outcome: Visit 4 and Any TB
Case definitions:
� Definite TB: Any culture positive in the presence of symptoms of pulmonary TB with or without positive smears for AFB.
� Probable TB: Culture negative, and at least one smear positive for AFB in the presence of symptoms of pulmonary TB.
� Possible TB: Negative smears, negative cultures for TB in the presence of symptoms of pulmonary TB with compatible chest X ray or clinical response to TB treatment (weight gain).
� No TB: Negative bacteriology and culture for TB and resolution of symptoms without TB treatment with or without treatment for another condition or an alternative diagnosis made.
Smearn=139
MDRPlus (Hain)n=121
LightCycler MTB
(Roche)n=130
Gene Xpert (Cepheid)
n=116
MGIT Culture
n=10745.2%
(34.4%;56.4%)53.5%
(41.4%; 65.3%)54.4%
(42.8; 65.5%)62.3%
(49.7%; 73.5%)73.8%
(62.8%; 82.5%)
100%(91.8%; 100%)
94.0%(82.4%; 98.4%)
94.1%(82.7%; 98.4%)
93.6%(81.4%; 98.3%)
100%(91.8%; 100%)
sensitivity
specificity
Available technology for Viral load in Tiered Laboratory Network
Tertiary Reference Laboratory Nucleic acid Testing strategies
Molecular methods: target or signal amplification
Roche Amplicor (manual and COBAS Amplicor, COBAS Ampliprep/ Amplicor)**
Roche Taqman, vs.1** and 2Abbott RealTime** (LCx obselete)
bDNA (K-PCR)
Nuclisens EASYQ (mini-MAG or EASYMAG)* version 2
In-house assays*
Biocentric assayRNA qualitative screening test
Second Tier May do NAT
Exavir Load
Primary Care POC*Sample collection: DBS
Technology Landscape for HIV viral load POCT
� Sudden and dramatic expansion of options� Specimen Processing has been the bottle-
neck previously� Technologies can be divided:
• In clinical validation• Close to clinical validation • Early development• Recognized Technologies that may in future be
applied to this use
Department of Molecular Medicine and Haematology
Category Technology General Principle
In early clinical validation LIAT (IQUUM) Multiplex real-time PCR
SAMBA (DRW) Isothermal amplification, Dipstick readout, plan for automation and visual read-out
HDA (Biohelix) Helicase dependent amplification
NAT technology that could be adapted
GeneXpert Cepheid Multiple targets, real-timeTB, MRSA
BD Max (Handylab) Real-time microfluidic PCRInfluenza
Enigma (Enigma diagnostics)
Twist Dx Recombinase polymerase amplification (RPA)
Early development EO-NAT HIV (wave 80)
Semi-Bio
North-western University p24
Eiken chemical
Inverness, Abbott etc
© IQuum, Inc. 2010
LIAT analyzerLiat TubeLiat Tube
1. Collect sample1. Collect sample
2. Scan barcode2. Scan barcode
3. Insert the tube3. Insert the tube
Get resultsGet results
Liat AnalyzerAssay chemistry: silica magnetic bead extraction & multiplex real-time PCR detection
Internal control strategy: RNA process control to minimize likelihood of a false negative readout
Time to result: 1hour
200ul plasma
Department of Molecular
Medicine and Haematology
Liat HIV Quant Assay (59 mins) compared to Abbott m2000 HIV Assay
� 93 HIV clinical samples tested
� 1 sample is Abbott assay low positive, Liat Assay negative
� 1 sample is Liat assay low positive, Abbott Assay negative
� For concordant samples, the viral load result had a 96.5% correlation coefficients between the 2 assays
� US based studies ongoing
y = 0.946x + 0.8302
R2 = 0.9652
1
2
3
4
5
6
7
8
1 2 3 4 5 6 7
Abbott HIV Viral Load [log10]
Liat
Vira
l Loa
d [lo
g10]
Early Field validation studies
Shugi Chen, 2010, IQUUM
LIAT
y = 0.9817x + 0.1187R2 = 0.9157
0
1
2
3
4
5
6
7
0 1 2 3 4 5 6 7
Reference Assay VL(log10)
Lia
t A
ssay
VL
(lo
g10)
�92% correlation with Abbott m2000 with 75 clinical specimens (clades A, B, C, and D) �Training took 5 minutes�Easy to use�Assay takes 60 minutes
�Fiscus, unpublished data 2010 (IAS, Vienna)
Department of Molecular
Medicine and Haematology
LIAT Summary
•• Turnaround time Turnaround time of 1 hour compared to several hours to days for existing assays
•• Full automation from sample to result for POCFull automation from sample to result for POC with no sample manipulation, reagent handling, or use of supporting laboratory equipment such as centrifuges, vortexes, pipettes, or computers
•• Able to accept POC compatible sample matricesAble to accept POC compatible sample matrices such as saliva and whole blood
•• Portable and battery operation Portable and battery operation to enable bedside or field use
•• Maintains sensitivity at low sample volumeMaintains sensitivity at low sample volume , ideal for pediatric patients
• Meets requirements of MSF round-table discussion document
In early clinical validation: SAMBA technology
• isothermal signal amplification• 2 hours from collection to result• Automation of front end extraction underway• Preliminary data: sensitivity <400 RNA copies/ml
Combination Strategy likely in SA
• Total Assay numbers in April 09-March 2010:• 3,190,668 CD4 assays (69 sites)
• 1,233,462 viral load assays (18 labs)
• 254,738 HIV DNA PCR assays
• Point –of- Care throughput:• CD4: 20 minutes per assay : 8-10 per day
• Viral load: 1-2 hours per assay: 4-8 per day
• Assuming 22 days per month X 12 months
• Total numbers of instruments for POC alone:• 1510 analyzers for CD4
• 704-1408 analyzers for viral load and DNA PCR
Likely Challenges for Implementation….
• Equipment Selection• Volume basi s
• Appropriate method validation• Study design• Phase I: Review of technology• Phase II: Laboratory based validation against gold
standard methodology : accuracy, precision etc• Phase III: Limited and expanded field validation
(small pilot sites)• Phase IV: Evaluation of clinical outcome of
interventionDepartment of Molecular Medicine and
Haematology
Continued…
� Appropriate internal quality control� External quality assurance programs� Instrument maintenance strategy� Good supply chain� Data retention and consolidation� Interpretation at clinic sites� Biohazard disposal� Regulatory environment
Acknowledgements
� Debbie Glencross and Lindi Coetzee� Lesley Scott: GeneXpert data� NHLS staff� POCWG� POCWG from ACTG� Clinic sites: R.Osih, F.Venter, I.Sanne� CDC in SA