Options for Lab monitoring Tobias F Rinke de

Options for Lab monitoring Tobias F. Rinke de Wit Pharm. Access International t. rinkedewit@pharmaccess. org TM

With continuous drop of HAART prices. . . Prices (US$/year) of a first-line antiretroviral Regimen in Uganda: 1998 -2001

…. lab monitoring is getting more expensive then HAART Number of tests Lab costs per patient ($) Prices per test year 1 year 2 com. CD 4 T-cells 5 3 Viral load 6 2 2 Hematology 6 Blood chemistry: 4 parameters Commercial prices year 1 year 2 25 15 6 125 50 25 100 150 50 600 2 0. 5 1 3 1 6 2 5 2 1 2 20 8 40 16 Blood chemistry: 9 parameters 2 0 1 2 18 0 36 0 HIV serology 1 0 5 10 5 0 10 0 TB chest X-ray, AFB 1 0 10 25 10 0 25 0 Pregnancy test 1 0 5 10 5 0 10 0 Hepatitis serology 1 0 10 236 0 65 20 872 0 Subtotal min. Minimum prices Clinton Foundation ARV prices in South Africa: 140$/year 268

Is Lab conditional for HAART? • Absence of Lab should not be a reason to refrain from HAART • Clinical facilities, (rapid) HIV test and HAARTDOT could be an option (Haiti)

What does lab add? (1) • Inclusion criteria for HAART ~ 70% clinical decision* ~ 15% CD 4 decision* ~ 8% viral load decision* • Clinical management of HAART ~ success/failure criteria ~ prevention OI’s ~ salvage/second line decisions * T. Spira, Forum for Collaborative HIV Research, Warsaw, 2003

What does lab add? (2) • Managing drug toxicities / adverse reactions during HAART • HIV resistance surveillance • Pharmacokinetics • Operational research • Infrastructure for future vaccine trials

What can be done ? • Use simplified laboratory algorithms • Negotiate bulk price reductions • Fix samples, transport to central lab and benefit from bulk test prices • Use cheaper, simpler lab technologies • Use generic reagents • Develop new lab technologies

Simplified Lab algorithms 50% cost reduction Algorithm: 100% HIV serology CD 4 Haematology WBC diff Chemistry Pregnancy HIV Viral Load Hepatitis B+C Syphilis Urine X-ray Physical examn pre-intake Y Y Y Y wk 2 wk 4 wk 6 wk 12 wk 24 wk 36 wk 48 Y Y Y Y Y Y Y Y Y wk 6 wk 12 wk 24 wk 36 wk 48 Y Y Y Y Algorithm: 50% HIV serology CD 4 Haematology WBC diff Chemistry Pregnancy HIV Viral Load Hepatitis B+C Syphilis Urine X-ray Physical examn pre-intake Y Y Y Y wk 2 wk 4 Y Y Y

Price reductions FACScount has internal QC External QC: in combination with blood fixatives and QC program (e. g. QASI, Health Canada) January 14, 2004: “Clinton Foundation Announces Landmark Agreement On Major Reduction In Price Of HIV/AIDS Lab Tests”

Stability of whole blood controls with fixatives Transfixed Status flow normal 4 0 C Immuno-trol Fluoro-trol low Fluoro-trol mid Fluoro-trol normal 0 2 4 220 C Days 6 8 10 0 2 4 Days 6 Transfixed 3700 C C 37 Status flow normal 10 -20% Transfix 10 days blood stabilization Immuno-trol Fluoro-trol low Fluoro-trol mid Fluoro-trol normal 0 2 4 Days 6 8 10

FACScount • • Simple to use, single platform Internal QC in place Maintenance service (~$5, 000/y) in place Medium throughput machine (~100 samples/d) Absolute count only (no CD 4% for pediatric) Cold-chain dependent No generic alternatives for expensive reagents Investment (ex VAT, transport) >$32, 000* * Can decrease to $28, 000; refurbished machine = $17, 500; 6 $ per test

Cheaper, simpler lab technologies • CD 4 ~ Cytospheres ~ Dynabeads ~ Trax? • Viral load ~ P 24 ~ Cavidi ~ Real time PCR ~ DBS

Cytospheres 100 L Whole Blood MONOCYTE BLOCKING Cyto. Sphere CD 4 T-CELL ASSAY CD 4 BINDING LYSE SOLUTION ROSETTE COUNT

Cytospheres CD 4 T-cell 2. 1 m

Difference from average mean CD 4 Learning curve with Cytospheres 500 1 -First set (n-8) 2 -Second set (n-8) 3 -Third set (n-8) 400 300 200 100 0 -100 1 2 Sample sets 3

Dynabeads

Dynabeads with light microscope

Dynabeads with fluorescent microscope

Difference from average mean CD 4 Cytospheres Dynabeads 1 - Cytospheres 2 - Cytospheres 3 - Dynabeads 4 - Dynabeads 600 400 CD 4 800 (n=8) CD 4 > 800 (n=16) 200 mean -200 -400 1 2 3 4 Error estimation increases with higher CD 4 T-cell counts Dynabeads underestimate at higher CD 4 T-cell counts: too many events (no dilution) eye fatigue; at lower CD 4 counts: no problems

Dynabeads / Cytospheres • • • Simple, quite cheap (3 -8$/ CD 4 test) Counts <700/ul are reproducible Labor intensive (<15 /day): no batch Subjective interpretation, eye fatigue Underestimation (Dynabeads) Overestimation (Cytospheres) Not applicable for % CD 4 (pediatric) Blood sample should be fresh (< 6 h) Cold storage reagents required (fridge) No external QC possible (blood stabilizers)

Dried blood spots for CD 4* ? • DBS-TRAx ELISA technology is not updated + supported • 38% of CD 4 data <200/ul were excluded from analysis • Source of CD 4 cannot be determined ~ CD 4 from T-cells ? ~ CD 4 from monocytes: bacterial infections (e. g. TB) ? ~ CD 4 from eosinophils: parasitic infections ? * Mwaba et al. , Lancet Nov. 2003

Viral load technologies • Nucleic acid based – NASBA • Protein based – Perkin Elmer ultrasensitive P 24 ~ Bio. Merieux Nuclisens – PCR ~ Roche Amplicor/monitor 1. 5 ~ Real time (Taq. Man) – b. DNA ~ Bayer – Real time NASBA ~ Primagen Rainbow – LCx ~ Abbott LCR + PCR • RT activity based – Cavidi RT viral load

Nucleic acid based viral load tests • • Expensive equipment (30, 000 - 60, 000$) Expensive kits (25 - 60$/test) Contamination problems (DNA > RNA) Complex, training required Equipment maintenance problems Lab discipline Several dedicated lab rooms

Alternative VL technologies • • Ultrasensitive P 24 ELISA Cavidi RT assay Primagen Rainbow assay Dried blood spots (DBS)

Ultra-sensitive P 24 ELISA

P 24 advantages • • Good linearity (500 – 6. 250. 000 fg/ml) Inexpensive: 6 - 8$/assay Possible for pediatric applications Good correlation with nucleic acid-based viral load assays

P 24 disadvantages • • • ELISA machine dependent Cold chain dependent Subtype dependence (D) ? Fragile method, learning curve Not evaluated for African samples (high Ig. G, P 24/a. P 24 complexes) • Boiling destroys epitopes risk of cross reactions • m. Abs not against denatured P 24 • Less sensitive than PCR-based technologies: – Load 10. 000 -100. 000: 87% P 24+ – Load <10. 000: 66% P 24+

Cavidi RT assay • HIV subtype independent: ~ no primers ~ no m. Abs, but 1 ml of plasma required • Active product improvement ~ (o/n) incubation at 33 o. C is shortened ~ Lower LDL (~1, 000 copies/ml) ~ Robust conversion factor established: 1 fg = 220 copies/ml • Equipment needed, not provided ~ rocking table, vacuum-pump, incubator, plate washer, plate reader, magnetic spinner • Kit, if delivered, unfrozen should be used within 1 week • No positive control in kit (biohazard) • Plasma should be 4 -6 hours old, or plasma should be stored at – 80 o. C

Primagen Retina HIV-1 Rainbow assay: real-time NASBA • Add amplification buffer to isolated nucleic acid • Heat 2 minutes 65ºC • Cool in 5 minutes to 41ºC • Add enzyme mix • Place in thermostated (41ºC) fluorimeter

Primagen Rainbow-DBS Blood drawn or via finger stick Blood on the filter • For service testing or external QC: • <20 tests: 50 euro • >20 tests: 35 -45 euro Transport to service testing lab by mail

Rainbow - Roche Amplicor vs Rainbow Load Panel n = 88, r = 0. 919 Rainbow log cps/ml 7 6 5 4 3 Rsq = 0. 9871 2 thru origin 2 3 4 5 Amplicor log cps/ml 6 7

Use generic reagents: Pan-leukogating strategy SSC total leukocytes (TLC) CD 45 Cheap reagents, but expensive equipment SSC CD 4+ monocytes CD 4+ T-cells CD 4 Glencross et al. , Cytometry 50, 69 -77, 2002

Use generic reagents: Mobile Cyflow

Cy. Flow • • Stand-alone: volumetric absolute and percent count Easy to transport, works on battery (mobile labs) Machine: 21, 000 $, test: 3 $ (generic reagents) 10% underestimation CD 4, strange for a no wash/no lyse method (T. Spira et al. , 2003) No maintenance network Installation, calibration fragile: electrode shift volumetric mistakes (blood viscosity) Software: user unfriendly Not (yet) multi-center evaluated, no WHO approval

New technolgies • Easycount ~ Image analyzer ~ Problems with monocytes • Apogee Flow System A 40 ~ To be evaluated • Guava ~ ~ Capillary flow cytometer, no sheath Fingerprick sample sufficient 35, 000$ equipment, 2$ reagents CD 4 is not core-business • Dipsticks ~ Semi-quantitative, both for CD 4 and VL ~ Cheap (1 -2 $)