AUTOMATION OF NAT SCREENING

Volume 41, April 2001 TRANSFUSION 483www.transfusion.org

Serologic screening procedures have substantiallyreduced the risk of transmission of blood-bornepathogens. However, there are still residual risks,due largely to the units of blood donated by newly

infected individuals during the preseroconversion windowperiod.1,2 To reduce the duration of the window period, NATis being instituted in Europe and the United States.3 Al-though these assays can be carried out by using a 96-wellformat for high throughput, the viral nucleic acid-purifica-tion process is generally slow and labor-intensive. Becauseof the lack of widely available automated testing equip-ment, NAT is used only in pools of samples at this time,3 butthe FDA is encouraging the eventual testing of single dona-tions.4 Thus, a rapid method of isolating viral nucleic acidis needed.

The process of extracting and purifying nucleic acidshas been complicated and labor-intensive in the past, be-cause of organic extraction and precipitation steps.5 Theseprocedures require repeated centrifugation and careful re-moval of supernatant, steps that are difficult if not impos-sible to automate. Despite claims for convenient “direct”methods, which omit organic extraction and washing steps,such methods have been found to suffer from interference

Automation of nucleic acid extraction for NATscreening of individual blood units

Dong-Hun Lee and Alfred M. Prince

BACKGROUND: Automation of NAT for single units ofblood is currently hampered by the labor-intensive stepsinvolved in the extraction of nucleic acids from samplesbefore the amplification procedures. A new method hasbeen developed for the automation of these steps usinghydrophilic polyvinylidene fluoride (PVDF) filter plates.STUDY DESIGN AND METHODS: Quantitative nucleicacid recoveries from sera containing HCV, HIV, HBV,HAV, and human parvovirus B19 and from 3H-labeledHCV RNA were determined in parallel by the semi-auto-mated PVDF method and a single-column method(Qiagen). Quantitative PCR was performed.RESULTS: Similar recoveries of HCV, HIV, and HBV(with silica beads) were observed with the PVDFmethod and with the Qiagen single-column method. Thesensitivity of the PVDF-based PCR assay for HCV, HIV,and HBV in serially diluted serum samples was alwayswithin two serial dilutions of that obtained when theQiagen single-column method was used in the same as-says. With the use of 3H-labeled HCV RNA, recoveriesof approximately 70 percent were found by both meth-ods.CONCLUSION: The PVDF method will permit full auto-mation of the simultaneous extraction of nucleic acidfrom sera containing HCV, HIV, and HBV. This procedurewill permit NAT screening of individual units of blood, willreplace the current screening of pools, and will achieveimproved blood safety with reduced labor and costs.

ABBREVIATIONS: B19 = parvovirus B19; GuSCN = guanidine

thiocyanate; NYBC = New York Blood Center; PVDF =

polyvinylidene fluoride; UTP = uridine triphosphate.

From the Laboratory of Virology, The Lindsley F. Kimball Re-

search Institute of The New York Blood Center; and the Depart-

ment of Pathology, New York University School of Medicine,

New York, New York.

Address reprint requests to: Alfred M. Prince, MD, Labora-

tory of Virology, The New York Blood Center, 310 East 67th

Street, New York, NY 10021; e-mail: aprince@nybc.org.

Supported in part by Grant No. AI47349 from the NIH and

Grant No. RO1 DK56406 from the National Institute of Diabetes

and Digestive and Kidney Diseases, NIH.

Received for publication April 13, 2000; revision received

September 4, 2000, and accepted September 29, 2000.

TRANSFUSION 2001;41:483-487.

T R A N S F U S I O N C O M P L I C A T I O N S

LEE AND PRINCE

484 TRANSFUSION Volume 41, April 2001 www.transfusion.org

by the inhibitors that are present in many sera.6-8 In ourexperience, such samples can give false-negative results onundiluted samples. Thus, the majority of existing methodsare not suitable for automated PCR, because of the require-ment for centrifugations or the unreliability of the method.Methods that employ hybridization of extracted nucleicacids to specific probes attached to magnetic beads havebeen reported.9,10 Such methods are, in theory, eligible forautomation, but they are likely to be costly.

It is the purpose of this study to overcome the abovelimitations in the purification of viral nucleic acids by pro-viding efficient and economical steps for the extraction ofnucleic acids from serum, plasma, or cells. This methodmust include a capture step in which the desired nucleicacid can be specifically or nonspecifically bound to a solidphase, which will permit inhibitors and unwanted compo-nents to be removed by washing. In addition, the methodshould not include centrifugation, as that step is difficultto incorporate into automated technology.

In the method to be presented, hydrophilic polyvinyl-idene fluoride (PVDF) membranes, together with SiO2 toenhance HBV extraction, are used to capture nucleic acids.These are incorporated in a 96-well plate format, whichpermits the use of a vacuum manifold for capturing andwashing. This method is suitable for use with laboratoryrobotics for the total automation of NAT screening of indi-vidual donations.

MATERIALS AND METHODSVirus stocksA stock of HCV was obtained from a chronically HCV-in-fected New York Blood Center blood donor ([NYBC] Donor#292). The HBV-infected plasma was taken from a chroni-cally infected chimpanzee. The HIV stock was a cell-freesupernatant from HIV-infected human peripheral bloodlymphocytes. The parvovirus B19 (B19) stock was obtainedfrom Anne-Marie Couroucé, MD (Paris, France). The HAVstock was obtained from Stan Lemon, MD (Galveston, TX)as supernatant from HAV-infected BS-C-1 (African greenmonkey kidney) cells.

Sample preparationSingle-column method. A viral RNA kit (QiaAmp, Qiagen,Chatsworth, CA) was used as the control purificationmethod for HIV, HAV, and HCV throughout this study, pro-viding a standard for the evaluation of the PVDF method forthese viruses. A blood kit (QiaAmp, Qiagen) was similarlyused as a standard for HBV DNA and B19 DNA.

PVDF/robotics procedures. A 0.22-µm hydrophilicPVDF membrane filter plate (Millipore, Bedford, MA) wasused for viral nucleic acid extraction in this study. Plasma(70 µL) was lysed with 250 µL of lysis buffer (5 M guanidinethiocyanate [GuSCN], 40 mM Tris-HCl, 20 mM EDTA, and1% Triton X-100 or GuSCN-based lysis buffer, such as AVL

buffer [Qiagen]) in a 96-well plate for 10 minutes at roomtemperature. ETOH (250 µL) was mixed with the lysate byautomated repeat pipetting (n = 10) (Biomek 2000,Beckman, Fullerton, CA). For HBV extraction, 20 µL of silicabead suspension (4 µm, 20% solids, Bangs Laboratories,Fishers, IN) and 250 µL of ETOH were mixed simulta-neously, and the mixture was incubated at room tempera-ture for 10 minutes before proceeding to the next step. Analiquot (300 µL) of above solution was transferred to thePVDF filter plate and vacuum-filtered at 600 mmHg. The re-mainder of the solution in the deep-well plate was trans-ferred to the PVDF filter plate to load a total 570 µL of thelysate/ETOH-filtered and vacuum. Washing buffer (300 µL:50 mM NaCl, 10 mM Tris-HCl, and 2 mM EDTA in 50% etha-nol or AW2 buffer [Qiagen]) was delivered to the plate andthen filtered at 675 mmHg for each of three washes. Thebottom of the plate was blotted onto filter paper (3mm,Whatmann) to remove drops of wash solution. Vacuumingat 675 mmHg was applied to the filter plate for 10 minutesto remove residual washing buffer, the bottom was againblotted, and the plate was air-dried for 10 minutes. Nu-clease-free H2O (50 µL) was added to the filter plate andvacuum-filtered at 600 mmHg to elute the samples into aU-bottom 96-well plate. Pipetting steps for reverse tran-scription and for the PCR steps were done by using the ro-botic capabilities of the Biomek 2000 pipetting station.

Quantitative PCR assayQuantitative PCR assays were carried out by using an au-tomated assay system (Ampli Sensor, Biotronics, Lowell,MA), in accordance with the manufacturer’s instructions.The AmpliSensor assay system monitors the amplificationefficiency of the PCR via a fluorescence resonance energytransfer-based detection scheme.11 The cDNA productswere reverse-transcribed from viral RNA. Then, the prod-ucts were amplified in an asymmetric manner to generatesingle-strand target DNA. These single-strand productswere reamplified in a semi-nested manner with a fluores-cent probe and primers. Fluorescence was measured byusing a fluorescence microplate reader (AmpliSensorMinilyzer, Biotronics). Quantitation is based on serial fluo-rescence measurements carried out between the 22nd and41st PCR cycles and on a computer-generated standardcurve derived from a dilution series of known quantities ofsynthetic RNA transcripts or from plasmid DNA molecules.

Reverse transcription. The reaction was carried out ina 20-µL volume containing 10 µL of PVDF eluate, 20 U ofM-MLV reverse transcriptase (Gibco BRL, Rockville, MD),12 U of RNAse inhibitor (RNasin, Promega, Madison, WI),50 mM Tris-HCl (pH 8.9), 75 mM KCl, 10 mM DTT, 1.5 mMMgCl2, 0.1 mM dNTP, and 0.1-µM reverse primers (HIVGAG: 5´-atgct ggtag ggcta tacat-3´, HCV 5´ UTR: 5´-gcgacccaac actac tcggc ta-3´). The reaction was run at 42°C for 45minutes and at 90°C for 2 minutes on a thermal cycler (AG9600 Cycler, Biotronics).

AUTOMATION OF NAT SCREENING

Volume 41, April 2001 TRANSFUSION 485www.transfusion.org

Asymmetric amplification. PCR master mix (4 µL) wasadded to the above cDNA reaction tube. The PCR mastermix contained 1 U of Taq polymerase (Promega), 50 mMTris-HCl (pH 8.7), 40 mM KCl, 1 mM DTT, 4 mM MgCl2, and0.9 µM of forward primers (HIV GAG: 5´-gaacc aaggg gaagtgacat agcag-3´, HCV 5´ UTR: 5´-gaaag cgtct agcca tggcg ttagta-3´; no additional reverse primers were added).12 HBV PCRassay was carried out in a 20-µL volume, using primers de-rived from the region encoding surface antigen (0.9 µMforward primer 5´-tgctc gtgtt acagg cgggg t-3´; and 0.1 µMreverse primer: 5´-gaggc atagc agcag gatga agag-3´).13 A 25-cycle amplification was run at 95°C for 30 seconds, 60°C for30 seconds, and 72°C for 45 seconds on a thermal cycler (AG9600).

Semi-nested amplification and detection. Amplisensorprimer duplex (4 µL; 1.5 ng/µL, Biotronics) was added to theabove PCR product mixture. One amplification cycle wasrun at 95°C for 30 seconds, 60°C for 30 seconds, and 72°Cfor 45 seconds on a thermal cycler (AG 9600) to establish abase reading for the fluorescence read-out system(Amplisensor Minilyzer); thereafter, assay readings werecarried out at every three PCR cycles until cycle 41.

HAV RT-PCR and B19 PCRHAV RNA was reverse-transcribed in a 20-µL reaction. TheHAV cDNA and B19 DNA were amplified as previously de-scribed, using equal amount of forward primers (HAV: 5´-gccgt ttgcc taggc tatag-3´; B19: 5´-aagtttgccg gaagt tcccg-3´)and reverse primers (HAV: 5´-ctcct acagc tccat gcta-3´; B19:5´-agcat cagga gctat acttc-3´) in 50-µL reactions. A 15-µLportion of the PCR product was run on a 5-percentpolyacryamide gel, stained with ethidium bromide, andvisualized under UV light.

In vitro transcription of synthetic 3H-labeled HCVRNA and 3H countingFull-length HCV-H strain (genotype 1b) was cloned into avector (pBluescript-SK, Stratagene, La Jolla, CA), which con-tains a T7 promoter upstream of the 5´ UTR. HCV RNA tran-

scripts were synthesized from the linearized cDNA tem-plates using a polymerase kit (MEGAscript T7, Ambion, Austin,TX).14 For labeling, uridine triphosphate (UTP) and 3H-UTPwas mixed at 17:1 ratio. Serial dilutions of labeled 3H-HCVRNA were spiked into human plasma lysate and immedi-ately subjected to Qiagen column and PVDF purifications.Untreated 3H-RNA counts were compared with those ofpurified samples to obtain percentage recoveries. All thereadings were generated on a liquid scintillation counter(Microbeta, Wallac, Gaithersburg, MD) using a scintillationfluor (Optiphase SuperMix, Wallac, Milton Keynes, UK).

Statistics

The PVDF and the Qiagen column methods were comparedby using the data from RNA/DNA quantities measured byquantitative PCR and by counting 3H by pairwise t test.

RESULTS

Serial dilutions of HCV plasma samples were extracted byusing the PVDF and Qiagen single-column methods andthen subjected to quantitative PCR analysis. The recoveryof HCV RNA isolated by the PVDF method was quantitativein all three dilutions and not significantly different from thatobtained by the Qiagen single-column method (Table 1). Toconfirm these results, the percentage of recovery was de-termined by using in vitro synthesized 3H-labeled HCVRNA. On the basis of 3H counting, HCV RNA recoveries av-eraged approximately 70 percent for both the PVDF and theQiagen single-column methods (Table 1). The PVDFmethod had a lower recovery than the Qiagen columnmethod only for the highest HCV RNA concentration tested.This contained approximately 1010 to 1011 RNA moleculesper mL, depending on the extraction volume; this concen-tration is unlikely to be encountered in clinical specimens.These results confirmed the PCR data and further demon-strated that hydrophilic PVDF is as effective a medium asthe Qiagen column for capturing HCV RNA when less than40 ng (8 × 109 molecules) was applied.

TABLE 1. Comparison of HCV RNA recovery with Qiagen column and PVDF methodsQuantitative PCR Percentage recoveries determined by 3H counting

RNA recovery (log RNA mol/mL) Percentage recovery

HCV* Qiagen HCV RNA† Qiagendilution Number column PVDF filter (molecules) Number‡ column PVDF filter

10–2 6 06.27 ± 0.05§ 6.21 ± 0.09 000.4 ng (8 × 107) 6/6 051.67 ± 10.84|| 60.17 ± 29.4210–3 6 5.24 ± 0.09 5.16 ± 0.07 04 ng (8 × 108). 6/5 75.00 ± 16.71 75.60 ± 13.9010–4 6 4.18 ± 0.16 4.04 ± 0.12 40 ng (8 × 109). 6/6 86.00 ± 11.52 70.17 ± 6.49¶

Average 18/17 070.89 ± 19.310 68.24 ± 19.36

* Human serum from an HCV patient (NYBC #292), diluted in human plasma.† 3H HCV RNA transcript generated from HCV-H strain.‡ Qiagen/PVDF.§ Quantity ± SD determined by PCR assay using the AmpliSensor system.|| Percentage recovery ± SD determined by 3H counting.¶ Significantly different from Qiagen (p<0.05).

LEE AND PRINCE

486 TRANSFUSION Volume 41, April 2001 www.transfusion.org

To determine whether the PVDF method is also effec-tive for other virus-extraction procedures, HIV (RNA), HBV(DNA), B19 (DNA), and HAV (RNA) serum samples weretested. HAV was included, despite the realization that thisvirus is not a major blood-borne pathogen. It can, however,be rarely transmitted by blood transfusion,15 and, moreimportant, it can be transmitted by pooled FVIII prepara-tions16; thus, HAV NAT screening of plasma used for themanufacture of FVIII would be desirable. Similarly, B19 wasincluded, as it is an important and frequent contaminantin pooled plasma and plasma-derivative preparations thatcould be eliminated by the testing of source plasma. Therecoveries of HAV RNA and B19 DNA isolated by the PVDFmethod were better than those obtained by use of theQiagen columns. Yields of amplicons were assessed onpolyacrylamide gels (Fig. 1).

The recoveries of HIV RNA isolated by the PVDF methodwere not significantly different from those obtained by theQiagen column method (Table 2). The recovery of HBV DNAwas lower than that with Qiagen column, but the addition

of 4 mg of silica beads to the sample lysate before vacuumfiltration significantly improved the HBV recovery, althoughthe Qiagen columns were slightly more sensitive (Table 2).The application of more than 4 mg of silica bead solutiondid not improve the recovery (data not shown). The addi-tion of silica beads did not interfere with extractions of HAVRNA, B19 DNA (Fig. 1), HIV RNA, or HCV RNA, except for aminor reduction seen at the 10–2 dilution (Table 2).

We also evaluated HBV DNA extraction using SDSbuffer with protease (Qiagen blood kit), because GuSCN-based lysis does not completely remove covalently boundprotein from the 5´ end of the L strand of HBV DNA.17,18

When SDS-based buffers were used in the PVDF method incombination with silica beads, the PVDF method was aseffective as the Qiagen blood kit method (data not shown).

The PVDF method, using 96-well hydrophilic PVDF fil-ter plates, was integrated into the pipetting station (Biomek2000) to test semi-automated high-throughput capability.The robotics were programmed to deliver and transfer liq-uids in the PVDF procedures for HCV extraction from plasmasamples, and HCV RNA capture, wash, and elution wereachieved by vacuum filtration. Chronically HCV-infectedchimpanzee plasma samples, prepared by the PVDF/robot-ics procedures, displayed the same viral loads as found bythe Qiagen column method in the PCR assay (Table 3). Al-though there were slight sample-to-sample variations in thequantities recovered by the two methods, the mean of thedata showed no difference. These results demonstrated thatthe semi-automated sample preparation has sensitivityequivalent to that of the Qiagen column method in the PCRassay. The method was technically feasible for high-throughput undiluted sample preparation for PCR.

DISCUSSIONThis study addressed the automation of a nucleic acid ex-traction procedure suitable for use in NAT screening of in-

dividual units of blood. NAT screening iscurrently carried out with pools, in partbecause no sample preparation meth-ods are available that are suitable forautomation and high-throughput test-ing. The testing of individual units,rather than pools, is likely to be particu-larly important for reducing the risk oftransmission of HBV and HIV. Theformer has a long window period withgradually increasing viral titers.4 Thelower-titer samples would be likely to bemissed by pool testing. HIV infectionshave a long intermediate phase duringwhich viral titers may be too low to bedetectable in pools.

Our data show that the PVDF filtermethod is suitable for single-unit NAT, in

Fig. 1. A 402-bp product from B19 DNA by PCR and a 237-bp

product were amplified from HAV RNA by RT-PCR. PCR prod-

ucts obtained using Qiagen column (Q), the PVDF filter (F),

and PVDF + Silica beads (FS) methods were compared for RNA

recovery from supernatant of HAV-infected BSC-1 cells and

for DNA recovery from B19-infected serum, both diluted in

normal human plasma.

TABLE 2. Comparison of HIV RNA, HCV RNA, and HBV DNA recoverieswith Qiagen column, PVDF, and PVDF/silica methods

Viral load (log DNA mol/mL)

Virus Dilution n Qiagen column PVDF PVDF/silica

HIV* 10–1 6 07.62 ± 0.10† 7.53 ± 0.09 7.52 ± 0.1110–2 6 6.58 ± 0.09 6.46 ± 0.14 6.47 ± 0.0810–3 6 5.35 ± 0.15 5.40 ± 0.14 5.25 ± 0.06

HCV‡ 10–1 6 7.42 ± 0.05 7.30 ±0.13 7.29 ± 0.1410–2 6 6.43 ± 0.05 6.25 ± 0.17 06.12 ± 0.22§10–3 6 5.38 ± 0.03 5.27 ± 0.20 5.20 ± 0.27

HBV|| 10–1 6 6.25 ± 0.06 05.79 ± 0.13§ 006.10 ± 0.09§¶10–2 6 5.24 ± 0.09 04.85 ± 0.23§ 05.05 ± 0.15§10–3 6 4.29 ± 0.05 03.96 ± 0.20§ 04.28 ± 0.15¶

* RNA was extracted from cell culture supernatant and diluted in human plasma.† Quantity ± SD determined by PCR assay using the AmpliSensor system.‡ Human serum from an HCV patient (NYBC #292), diluted in human plasma.§ Significantly different from Qiagen (p<0.05).|| HBV 75-546 challenge inoculum diluted in normal human plasma.¶ Significantly different from PVDF (p<0.05).

AUTOMATION OF NAT SCREENING

Volume 41, April 2001 TRANSFUSION 487www.transfusion.org

that it can be readily adapted to total robotic automation.The PVDF method was found to have a sensitivity equal toor in some cases slightly less than that of one of the mostsensitive extraction methods,19,20 Qiagen columns. In allcases, the sensitivities were within two serial dilutions ofeach other. Further optimization is likely to eliminate anydifferences.

It is important that the PVDF method has been shownto be suitable for extracting both DNA viruses such as HBVand parvovirus B19 and RNA viruses such as HIV, HCV, andHAV with high sensitivity and using the same procedure.Thus, the method is suitable for use before a multiplex NATprocedures to detect both DNA and RNA viruses.

REFERENCES01. Busch MP, Stramer SL, Kleinman SH. Evolving applications

of nucleic acid amplification assays for prevention of virus

transmission by blood components and derivatives. In:

Garratty G, ed. Applications of molecular biology to blood

transfusion medicine. Bethesda: American Association of

Blood Banks, 1997:123-76.

02. Schreiber GB, Busch MP, Kleinman SH, Korelitz JJ. The risk

of transfusion-transmitted viral infections. The Retrovirus

Epidemiology Donor Study. N Engl J Med 1996;334:1685-90.

03. Wilkinson S, Lipton KS. NAT implementation. AABB Asso-

ciation Bulletin #99-3. Bethesda: AABB, February 8, 1999.

04. Busch MP, Kleinman SH, Jackson B, Stramer SL, Hewlett I.

Committee Report. Nucleic acid amplification testing of

blood donors for transfusion-transmitted infectious dis-

eases: report of the Interorganizational Task Force on

Nucleic Acid Amplification Testing of Blood Donors. Trans-

fusion 2000;40:143-59.

05. Chomczynski P, Sacchi N. Single step method of RNA isola-

tion by acid guanidinium thiocyanate-phenol-chloroform

extraction. Anal Biochem 1987;162:156-9.

06. Ali N, Jameel S. Direct detection of hepatitis C virus RNA in

serum by reverse transcription PCR. Biotechniques

1993;15:40-2.

07. Ravaggi A, Prim D, Cariani E. Direct PCR amplification of

HCV PCR from human serum. PCR Methods Appl

1992;1:291-2.

08. Lai J, Prince AM, Wolf L, Andrus L. A simplified method for

PCR detection of hepatitis C virus RNA from human serum.

PCR Methods Appl 1994;3:308-9.

09. Miyachi H, Masukawa A, Ohshima T, et al. Automated spe-

cific capture of hepatitis C virus RNA with probes and para-

magnetic particle separation. J Clin Microbiol 2000;38:18-21.

10. Beaulieux F, See DM, Leparc-Goffart I, et al. Use of mag-

netic beads versus guanidium thiocyanate-phenol-chloro-

form RNA extraction followed by polymerase chain reac-

tion for the rapid, sensitive detection of enterovirus RNA.

Res Virol 1997;148:11-5.

11. Wang CN, Wu KY, Wang HT. Quantitative PCR using the

AmpliSensor assay. In: Dieffenbach CW, Dveksler GS, eds.

PCR primer: a laboratory manual. Cold Spring Harbor, NY:

Cold Spring Harbor Laboratory Press, 1995:193-202.

12. Gretch D, Lee W, Corey L. Use of aminotransferase, hepati-

tis C antibody, and hepatitis C polymerase chain reaction

RNA assays to establish the diagnosis of hepatitis C virus

infection in a diagnostic virology laboratory. J Clin

Microbiol 1992;30:2145-9.

13. Kaneko S, Feinstone SM, Miller RH. Rapid and sensitive

method for the detection of serum hepatitis B virus DNA

using the polymerase chain reaction technique. J Clin

Microbiol 1989;27:1930-3.

14. Schenborn ET, Mierendorf RC Jr. A novel transcription

property of SP6 and T7 RNA polymerase: dependence on

template structure. Nucleic Acids Res 1985;13:6223-36.

15. Lee KK, Vargo LR, Le CT, Fernando L. Transfusion-acquired

hepatitis A outbreak from fresh frozen plasma in a neonatal

intensive care unit. Pediatric Infect Dis J 1992;11:122-3.

16. Vermylen J, Peerlinck K. Review of the hepatitis A epidemics

in hemophiliacs in Europe. Vox Sang 1994;67(Suppl 4):8-11.

17. Boom R, Sol CJ, Heijtink R, et al. Rapid purification of hepa-

titis B virus DNA from serum. J Clin Microbiol 1991:29;1804-11.

18. Heermann KH, Gerlich WH, Chudy M, et al. Quantitative

detection of hepatitis B virus DNA in two international ref-

erence plasma preparations. The Eurohep Pathobiology

Group. J Clin Microbiol 1999:37;68-73.

19. Fransen K, Mortier D, Heyndrickx L, et al. Isolation of HIV-1

RNA from plasma: evaluation of seven different methods of

extraction (part two). J Virol Methods 1998;76:153-7.

20. Shafer RW, Levee DJ, Winters MA, et al. Comparison of

QIAamp HCV kit spin column, silica beads, and phenol

chloroform for recovering human immunodeficiency virus

type 1 RNA from plasma. J Clin Microbiol 1997;35:520-2.

TABLE 3. Comparison of HCV RNA recovery fromchronically infected chimpanzees by Qiagen column

and PVDF/robotic methodsRNA recovery (log RNA mol/mL)

Chimp Log Qiagen column PVDF/robotic

167 98-190 *5.67* 6.00173 98-270 6.48 6.58173 98-273 8.08 8.02208 98-257 7.45 7.36208 98-337 6.98 6.72214 98-315 5.00 5.47215 98-278 7.54 7.08215 98-301 6.53 6.73215 98-334 7.86 7.01285 98-254 7.13 7.08285 98-309 6.94 7.01

Mean ± SD 6.88 ± 0.92 6.82 ± 0.67

* Quantity determined by PCR assay using the AmpliSensor sys-tem.