Journal of Virological Methods 98 (2001) 9–16

Comparison of capillary electrophoresis sequencing with thenew CEQ 2000 DNA Analysis System to conventional gel

based systems for HIV drug resistance analysis 1

Patrick Merel*, 1, Isabelle Pellegrin1, Isabele Garrigue, Anne Caumont,Marie-Helene Schrive, Valerie Birac, Pascal Bonot, Herve Fleury

Laboratoire de Virologie, Centre Hospitalier Regional et Uni�ersite Victor Segalen, Bordeaux, France

Received 27 February 2001; received in revised form 23 May 2001; accepted 24 May 2001

Abstract

To date the majority of sequencing technologies have been based on use of gel plates. In this study sequencing bycapillary electrophoresis for HIV-1 genotyping on the CEQ 2000 sequencer (Beckman Coulter Inc.) has beeninvestigated and compared to an ‘in house’ protocol on the Prism-377 sequencer (Applied Biosystems) and to theHIV-1 TruGene kit (Visible Genetics Inc.), two gel plate-based systems. Plasma from 20 HAART-treated patientswith virological failure were analyzed for protease (PR) and reverse transcriptase (RT) genes. A total of 520 RTcodons (26/patient) and 360 PR codons (18/patient) related to antiretroviral drug resistance were evaluated. Theoverall agreement between CEQ 2000 and Prism-377 results was 100% for the RT and PR primary and secondarymutations. The overall agreement between CEQ 2000 and TruGene was 100% for primary and�97% for secondarymutations. Discrepant results would have never led to errors in genotype interpretation. Performances for a 24patients/week/one technician genotyping throughput were analyzed. For Prism-377, TruGene and CEQ 2000, manualprocessing required 5, 4 and 2,5 days, sequence data analysis needed additional 3, 1 and 2 days and cost/patientwas�49, 214 and 39 $, respectively. The CEQ 2000 sequencer offers a reliable alternative for fast and cost effectiveHIV drug resistance analysis. © 2001 Elsevier Science B.V. All rights reserved.

Keywords: HIV-1 genotyping; Capillary electrophoresis; Sequencing; Cost-effectiveness

www.elsevier.com/locate/jviromet

1. Introduction

Evidence concerning the clinical use of drugresistance testing is accumulating from retrospec-tive (Deeks et al., 1999; Lorenzi et al., 1999;Zolopa et al., 1999) and prospective (Baxter et al.,1999, Durant et al., 1999) studies. The studiessuggest that genotype data added to the knowl-edge of a patient’s drug history may improve the

* Corresponding author. Laboratoire de Virologie, HopitalPellegrin, Place Amelie Raba-Leon, 33076 Bordeaux Cedex,France. Tel.: +33-5-567-98723; fax: +33-5-567-95673.

E-mail address: patrick.merel@chu-bordeaux.fr (P. Merel).1 The two first authors contributed equally to this study.

0166-0934/01/$ - see front matter © 2001 Elsevier Science B.V. All rights reserved.

PII: S 0166 -0934 (01 )00338 -X

P. Merel1 et al. / Journal of Virological Methods 98 (2001) 9–1610

ability of clinicians to select a regimen for thosewho have been already exposed to antiretroviraldrugs. Indeed, the predictive value of drug resis-tance profile on the subsequent response to antivi-ral drugs, as well as their use as an early markerof virological failure, before a rebound in the viralload occurs has been reported (Deeks et al., 1998;Harrigan et al., 1998; Lanier et al., 1998). Sincethe value and use of HIV genotyping as a clinicaltool will continue to increase, understanding ofthe technologies and applications of the differentavailable systems will become a necessity for in-vestigators, clinicians and physicians. Sequencingof DNA is considered the ‘gold standard’ proce-dure for HIV genotyping because it identifies notonly point mutations and polymorphisms, butalso insertions, deletions, and frame shifts. At atime when increasing number of commercial HIVdrug resistance genotypic tests is available, it be-comes critical not only to evaluate their accuracyand reproducibility, but also their cost-effective-ness. The new capillary sequencer from BeckmanCoulter Inc. (BCI, Fullerton, CA), the CEQ 2000DNA Analysis System was compared with two gelplate-based systems available for HIV drug resis-tance analysis: the Prism-377 sequencer from Ap-plied Biosystems (AB, Foster City, CA) and theHIV-1 TruGene sequencing kit from Visible Ge-netics Inc (VGI, Toronto, Ontario, Canada).

2. Material and methods

2.1. Patients and samples

Twenty plasma samples (stored at −80 °C)from 20 HIV-1 infected patients were retrospec-tively analyzed for RT and PR genotypes. Allpatients were followed-up at 3 months intervalsfor HIV-1 plasma viral load and had previousexposure to at least one protease inhibitor (PI)drug and at least two nucleoside reverse transcrip-tase inhibitor (NRTI) drugs. RT and PR geno-types were performed when virological failureoccured (HIV-1 RNA�� 1000 copies/ml). Eachgenotyping procedure was undertaken by a well-trained technician, up to sequence confirmation.A single expert analyzed the samples with the

CEQ 2000 sequencer in a blinded fashion and theresults were compared to the original genotypetest results derived from VGI or AB sequencersystems. The classification of the RT and PRresistance–associated primary and secondary mu-tations identified by genotyping into primary, sec-ondary mutations associated or not withdecreased drug sensitivity, was established accord-ing to the consensus statement on antiretroviraldrug-resistance (Hirsch et al., 2000).

2.2. HIV-1 genotyping

Plasma HIV-1 RNA was used for sequencingthe major part of the RT gene and the entire PRgene. When using the BCI-CEQ 2000 or the AB-Prism 377 sequencers, HIV-1 RNA was purifiedfrom 1 ml ultracentrifugated (23 500×g) plasmausing the High Pure Viral RNA Purification kit(Roche Diagnostics, Mannheim, Germany). Whenusing the TruGene HIV-1 Genotyping Kit, HIV-1RNA was purified with the QIAamp Viral RNA(QIAGEN, Hilden, Germany).

2.2.1. Sequencing procedure with the CEQ 2000sequencer

The CEQ 2000 DNA Analysis System is a fullyautomated capillary electrophoresis DNA se-quencing system. With a pre-assembled array ofeight capillaries and a 96-well format compatibleplatform, the CEQ 2000 allows high-throughputsequencing with full automation. The need for gelcasting (automatic gel pumping into capillariesfrom pre-filled gel cartridge), plate washing andsample loading is eliminated. Use of the CEQ2000 does not need modification of common se-quencing procedure except for the sequencing dyeterminator reagents. Briefly, HIV-1 RNA was am-plified by one-step RT-PCR (Reverse Transcrip-tion and Polymerase Chain Reaction) using theTitan One Tube RT-PCR Kit (Roche Diagnos-tics, Mannheim, Germany) followed by a nestedPCR with AmpliTaq Gold (AB). Operations wereconducted into microtiterplates. Based on theHIV-1, B–FR.LAI sequence, the primers MJ35� - AGTAGGACCTACACCTGTCAAC - 3�/MJ45�-CTGTTAGTGCTTTG-GTTCCTCT-3� (bases2480–2501, and 3399–3420) and A(35) 5�-TTG-

P. Merel1 et al. / Journal of Virological Methods 98 (2001) 9–16 11

GTTGCACTTTAAATTTTCCCATTAGTCCT-ATT-3�/NE1(35) 5�-CCTACTAACTT-CTGTAT-GTCATTGACAGTCCAGCT-3� (bases 2530–2560, and 3300–3334) were used for the reversetranscriptase gene RT-PCR and nested PCR, re-spectively. The primers 5�PROT1 5�-TAATTTTT-TAGGGAAGATCTGGCCTCC-3�/3�PROT1 5�-GCAAATACT-GGAGTATTGTATGGATTTT-CAGG-3� (bases 2082–2108, and 2703–2734) and5�PROT2 5�-TCAGAGCAGACCAGAGCCAA-CAGCCCC-3�/3�PROT2 5�-AATGCTTT-TATT-TTTTCTTCTGTCAATGGC-3� (bases 2136–2162, and 2621–2650) were used for the proteasegene RT-PCR and nested PCR, respectively. Af-ter nested PCR, PCR fragments (805 base pairsfor the reverse transcriptase fragment (positions2530-3335) and 505 base pairs for the proteasefragment (positions 2136-2640)) were purified byfiltration onto microtiterplate with the MultiS-creen PCR product purification kit (referenceMANU03010, Millipore, Bedford, MA). PCRproduct purity and concentration (�200 ng) wereestimated by visualization after agarose gel elec-trophoresis. Sequencing reactions were carried outwith primers A(35) and NE1(35) for sense andantisense RT sequencing and primers 5�PROT2and 3�PROT2 for sense and antisense PR sequenc-ing, according to the instructions included in theBeckman Coulter CEQ Dye-Terminator Cycle Se-quencing kit (reference 608000). This kit includesreagents for cycle sequencing by conventional oneprimer/one tube Dye Terminators reaction, withBCI proprietary dyes. Thermal cycling occurredon a GeneAmp PCR System 9700 (AB) and in-cluded the following steps: 96 °C for 10 s, 48 °Cfor 5 s, and 60 °C for 2 min. This cycle was run30 times. DNA sequencing products were purifiedby centrifugation into microtiterplate (MultiS-creen Dye Terminators removal kit, referenceMAHVN4550, Millipore). Deionized formamide(25 �l) was added to each sample before transferto the CEQ 2000 DNA Analysis System. Sequenc-ing run conditions were carried out automaticallyby the CEQ 2000 for eight-sample serie. First, thesamples were denatured 120 s at 90 °C, then aftera pause of 1 min, they were injected at 2 kV for 60s. Separation was done at 7 kV during 180 minfor RT gene sequencing and during 120 min for

the PR gene. Under these conditions, length ofsequencing fragments were�650 bp for RT genesequences and�300 bp for Prot gene sequencesin both directions. At run completion, the CEQ2000 software performed sequence analysis auto-matically then the sequence was exported to third-party DNA analysis specialized software.Sequencher (Gene Codes Corp, Ann Arbor, MI)for Windows (version 4.05) was used for sequencealignment on the HIV-1LAI RT and PR genes.

2.2.2. Sequencing procedure with the Prism-377sequencer

RT-PCR and nested-PCR steps were carriedout as described for the CEQ 2000 procedure.Nested-PCR products were purified with Mi-croSpin S-400 HR columns (Amersham Pharma-cia Biotech, Uppsala, Sweden). After agarose gelelectrophoresis control, the purified products weresequenced with the same primer pairs (sense+an-tisense) used during the CEQ 2000 procedure. TheAB-Prism Big Dye-Terminator Cycle SequencingKit was used, with the following cycle conditions96 °C for 10 s, 50 °C for 5 s, and 60 °C for 4min. This cycle was run 25 times. Sequence prod-ucts were then purified by Alcohol precipitationprior to gel loading on a 36 lanes Prism-377automatic sequencing System. Electrophoresis runwas performed at 3000 V during 3.5 h at 49 °C.Under these conditions, the length of sequencingfragments were�650 bp. After the run, sequenceswere aligned on the HIV-1LAI RT gene and theconsensus-B PR gene using the Sequence Analysisthen Sequence Navigator softwares (AB). Assign-ment of RT and PR resistance-associated muta-tions was performed as described in the CEQ2000 sequencing procedure.

2.2.3. Sequencing procedure with the TruGeneHIV-1 genotyping kit

The TRUGENE HIV-1 assay is used with theOpenGene automated DNA sequencer system(VGI). After RNA extraction, the first step of thekit involves RT-PCR of a single 1.3-kb fragmentof the HIV-1 genome that covers the whole PRgene and the major part of the RT gene. Thesecond step involves CLIP™ reactions for theentire protease gene and codons 35–244 of the

P. Merel1 et al. / Journal of Virological Methods 98 (2001) 9–1612

RT gene sequencing. The 1.3 kb fragment fromthe first step is split into three amplicons desig-nated Protease, RT beginning and RT middle.During dye-primer sequencing reactions, the for-ward and reverse primers are labeled with twodifferent fluorescent dyes, and mixed in a singletube. For each three amplicons, four reactionsneed to be setup per dideoxynucleotide (ddNTP).Thus, 12 sequencing products are analyzed in 12separate wells on a gel (MicroCell cassette, VGI)for rapid electrophoresis on a single Long-Read Tower sequencer (VGI). The OpenGeneVGI software combines the results across the fourlanes to generate the composite sequence. Theresulting assays are base called with GeneObjects(VGI), a software which compared the HIV-B-LAV1 reference sequence to the generated se-quences and to a resistance mutations database(GeneLibrarian, VGI). Mutations contained inthe sample sequence are finally reported by Ge-neObjects.

3. Results

3.1. Concordance among the three tests

The results obtained with the three methods forRT and PR genes mutations showed 100% agree-ment for primary mutations and more than 97%agreement for secondary mutations (Table 1).Disagreements, additional or missing mutationsfound by the CEQ 2000 were always confirmed byretesting samples from RNA extraction to se-quence reactions. Globally, of 880 investigatedcodons, 10 discrepancies were found, all locatedat secondary mutation sites, with 5/10 found in amixture viral population form (three with CEQ2000, two with TruGene). Despite these minordiscrepancies, these three genotyping methodsshowed a similar capacity to identify clinicallyrelevant mutations related to patients’currentdrug regimens. Discrepant results would havenever led to a change in the genotype test resultinterpretation.

Additionally, polymorphism codon informa-tions have been collected. Considering the three

procedures, a mean of 7.5 and 22.75 polymor-phism mutations were found in the PR and RTgenes, respectively. Mean of differences in thenumber of polymorphism mutations between thethree methods did not exceed 0.6 in the PR geneand 1.7 in the RT gene.

3.2. Methods time requirement

The time required by one technician for geno-typing PR and RT genes of 24 patients (Table 2)was compared for the three procedures. The CEQ2000 allowed the procedure to be completed in 2.5days of manual processing, followed by 2 days ofsequence data analysis. This time requirement wasobtained firstly because of rapid PCR productpurification with the Multiscreen PCR microtiter-plate (�15 min for 96 samples) and rapid dyeterminator removal procedure with the Multis-creen DT microtiterplate. Secondly, because ofthe CEQ 2000 automation capability, eight se-quence results can be analyzed as soon asthey are completed (every 2–3 h), while the se-quencer is running the following samples. Indeed,the system is setup in less than 10 min for 96sequences (=24 patient×one RT gene×one PRgene×one sense sequence×one antisense se-quence). Then, with gel injection carried out auto-matically, sample denaturation and sampleloading, 96 sequences were run in 33 h withunattended operation.

On the other hand, use of a 36 lanes Prism-377sequencer required 5 days of manual processingfollowed by 3 days of data analysis. Mainly be-cause Prism-377 users do not use traditional mi-crotiterplate options for sample manipulation, butalso because three gels need to be run for 96sequences electrophoresis. With a setup time re-quirement of almost 3 h per gel, and three runs of3.5 h each, this type of system was more timeconsuming compared to the capillary elec-trophoresis system.

The TruGene HIV genotyping kit and se-quencer provided a simple and rapid system forHIV genotyping. Each gel (16 lanes) prepared in apremade cassette, was setup within 5 min, andelectrophoresis completed within 50 min. Consid-ering a 24 patient batch, this procedure required

P. Merel1 et al. / Journal of Virological Methods 98 (2001) 9–16 13

Tab

le1

Gen

otyp

ing

assa

ysre

sult

(RT

and

PR

gene

sfr

om20

HIV

-1tr

eate

dpa

tien

ts)

conc

orda

nce

betw

een

capi

llary

sequ

enci

ngw

ith

the

CE

Q20

00an

dge

lba

sed

plat

ese

quen

cing

wit

hth

eP

rism

-377

and

the

Tru

Gen

eH

IVki

t

CE

Q20

00

Num

ber

ofR

Tan

dP

Rre

sist

ance

mut

atio

nsw

ith

the

follo

win

gte

stre

sist

ance

assa

yre

sult

s/C

EQ

2000

resu

ltco

ncor

danc

e

PR

(n=

20)

RT

(n=

20)

CSe

cond

ary

mut

atio

ns,

nine

Pri

mar

ym

utat

ions

,se

ven

test

ed/g

enot

ype,

Pri

mar

ym

utat

ions

,17

test

ed/g

enot

ype,

Seco

ndar

ym

utat

ions

,11

test

ed/g

enot

ype,

Cn=

340

test

ed/g

enot

ype,

n=

180

n=

140

n=

220

−/−

+/−

−/+

+/+

−/−

+/−

−/+

Gen

otyp

ing

assa

ys+

/++

/+−

/−+

/−−

/+−

/−+

/−−

/++

/+

Tru

Gen

e1

2811

20

0P

rim

ary

mut

atio

ns61

127

90

013

50

3a0.

9861

152

424b

3c0.

97Se

cond

ary

mut

atio

ns

Pri

sm- 3

771

2511

50

0P

rim

ary

mut

atio

ns64

127

60

013

30

01

6815

20

01

47Se

cond

ary

mut

atio

ns

+,

mut

atio

nde

tect

ed;−

,m

utat

ion

not

dete

cted

.W

T/M

,w

ild-t

ype

and

mut

ant

vira

lm

ixtu

re.

a ,in

clud

ed2

WT

/Mde

tect

edw

ith

CE

Q20

00.

b,

incl

uded

2W

T/M

dete

cted

wit

hV

GI.

c ,in

clud

ed1

WT

/Mde

tect

edw

ithC

EQ

2000

,C

,co

ncor

danc

eis

defin

edas

the

num

ber

ofte

sted

mut

atio

nsw

ith

conc

orda

ntre

sult/(

num

ber

ofte

sted

mut

atio

nsw

ith

conc

orda

ntre

sult+

num

ber

ofte

sted

mut

atio

nsw

ith

disc

orda

ntre

sult

s).

Cla

ssifi

cati

onof

the

RT

and

PR

mut

atio

nsin

topr

imar

yan

dse

cond

ary,

asso

ciat

edor

not

wit

hde

crea

sed

drug

sens

itiv

ity

was

esta

blis

hed

acco

rdin

gto

the

cons

ensu

sst

atem

ent

onan

tire

trov

iral

drug

-res

ista

nce

(9).

Pri

mar

yR

Tm

utat

ions

incl

uded

K65

R,

T69

D/N

,in

sert

ions

68–6

9,K

70R

,L

74V

,V

75I/

T,

V75

M/S

/A/T

,L

100I

,K

103N

,V

106A

,V

108I

,Q

151M

,Y

181C

,M

184V

,Y

188C

/H/L

,19

0A/S

,T

215Y

/F.

Seco

ndar

yR

Tm

utat

ions

incl

uded

M41

L,

A62

V,

D67

N/E

/S,

F77

L,

K10

1E,

F11

6Y,

L21

0W,

219Q

/E,

P23

6L.

Pri

mar

yP

Rm

utat

ions

incl

uded

D30

N,

M46

I/L

,G

48V

,I5

0V,

V82

A/F

/S/T

,I8

4V,

L90

M.

Seco

ndar

yP

Rm

utat

ions

incl

uded

L10

I,K

20R

,L

24I,

V32

I,M

36I,

I47V

,I5

4V/L

/T,

A71

V/T

,G

73S,

V77

I,N

88D

/S/T

.

P. Merel1 et al. / Journal of Virological Methods 98 (2001) 9–1614

18 gels to prepare. Finally, the VGI kit andsequencer combination performed the completeprocedure in 5 days.

3.3. Methods cost

The cost of one RT and PR sequence reactionwas 39, 49 and �$ 214 (French list price, fullvolume reactions) per patient, for CEQ 2000,Prism-377 and TruGene Kit, respectively. Thisincludes every necessary reagent, from RNA extra-tion to sequence analysis, but technican salary andannual service contract cost for the sequencers. Itshould be noted that the price calculated for theVGI assay included a reagents rental program witha sequencer at laboratory disposal whereas inFrance, BCI and AB sequencer prices are esti-mated at $ 91 202 and $ 118 497 respectively.

3.4. Methods software analysis

Whatever the procedure used, biocomputingtime necessary for sequence analysis was an impor-tant element in the genotyping process. Among the

three approaches tested, VGI was the only onewith dedicated software (GeneObjects) for se-quence confirmation and drug resistance mutationassignement. It allowed data analysis and reportoutput in 10 min/patient.

Transfer of data to a main LIMS required anadditional 10 min/patient: 8 h of computing workwere also necessary for 24 patient batch. Whenusing the CEQ 2000 with Sequencher for Win-dows, the process of data required 30 min/patientfor RT and PR sequence confirmation. Consider-ing the time to transfer these data to a LIMS, thecomputing process of data issued from a CEQ2000 was of 2 days (24×40 min). With a gel platebased sequencer such as the Prism-377, sequenceanalysis was made in two steps. After a gel run, thegel lanes were tracked and assembled with Se-quence Analysis software (AB), and then the se-quences were confirmed with Sequence Navigatorsoftware (AB): almost 1 h was needed per patient,thus bringing computing total time requirement toalmost 3 days (8×3×1 h). Finally, whateversoftware was used with the CEQ 2000 or thePrism-377, drug resistances were not assigned

Table 2Cost-effectiveness evaluation of the three HIV genotyping procedures

Protocol andSequencer type Cost/patient Cost (sequencer)Turn around time (days/24 patients)(reagents)technology(Company)

Manual Sequence Totalanalysisoperations

In house protocolCEQ 2000 (BeckmanCoulter Inc.)

91 202 $Capillary 2.5 2 4.5 39 $electrophoresisEight capillaryarray

In house protocolPrism-377 (AppliedBiosystems)

49 $835 118 497 $Gel basedsequencer36 Lanes

TruGene HIV-1OpenGene (VisibleGenetics Inc.) kit

514 Reagents rentalGel based 214 $sequencer16 lanes

P. Merel1 et al. / Journal of Virological Methods 98 (2001) 9–16 15

automatically. This can be done manually by theexpert, or by using web-based programs. HIV-1RT and PR Mutation Search Engine for Queries(HIV-SEQ) software from the Stanford HIV RTand Protease Sequence Database web site (http://hivdb.stanford.edu/hiv/programs.htm) or the An-tiretroviral Drug Resistance Analysis Tool(ADRA) from the Los Alamos National Labora-tory web site (http://hiv-web.lanl.gov/ADRA/adra.html) are of a great help to the expert. Theycan both import information from the CEQ 2000or the Prism-377 text or FASTA formatted se-quence files and generate a patient’s report.

4. Discussion

HIV drug resistance assays (genotyping andphenotyping) are now available commercially andtheir use is expanding rapidly in the clinical set-ting. This study describes the analytical perfor-mance and cost-effectiveness of a new capillaryarray sequencer, the CEQ 2000, using an ‘in-house’ protocol compared to the standardizedTruGene HIV-1 test and the Prism-377 ‘in-house’assay. These results, obtained on HIV-1 B sub-type isolates, have showed a similar analyticalperformance of the three systems for identificationof insertions, deletions and frame shifts implicatedin resistance to specific drugs. In contrast to otherstudies (Collin et al., 2000; Hoover et al., 2000;Wilson et al., 2000), less discrepancies were foundbetween results obtained by capillary sequencingand gel plate sequencing than when comparingresults from gel plate based systems. Discordanceswere found only between the CEQ 2000 and theTruGene procedure. Most of these discrepanciescorresponded to viral mixtures detected by one ofthe two tests only. It is well known that thesesystems identify mutations present only in thepredominant viral strains, which means that mu-tations corresponding to drugs used previously ormutations present in minor quasispecies, ie, �30% of a viral population, may not be recognized(Schuurman et al., 1999). The CEQ 2000 capacityto detect such minor populations is under investi-gation. Nevertheless, the Prism-377 is actuallyconsidered as the ‘gold standard’ in terms of

sequencer apparatus. A 100% result concordancebetween this sequencer and the CEQ 2000 wasshown and it is considered that the CEQ 2000 is areliable new alternative for HIV genotyping inclinical environment.

Aside from performance, cost consideration isimportant when implementing HIV genotypingtesting strategies. Indeed, conventional DNA se-quencing requires expensive equipment, to whichthe cost of hiring skilled and trained techniciansmust be added. However, economic analysis ofHIV drug resistance testing cost-effectivenessshowed that additional costs of genotyping wouldbe offset by saving in the cost of drugs (Simpson,2000), notably by avoiding prescription of inac-tive antiretroviral agents. The CEQ 2000 8-capil-lary array sequencer allows the design of amicrotiterplate compatible procedure for HIVgenotyping. From RNA to sequence interpreta-tion we describe a procedure that allows onetechnician to process a 24 patients batch in lessthan a week. The case of capillary electrophoresistechnology associated with the practability of thisinstrument minimizes errors due to sample han-dling and the overall processing time. Mainly forthese reasons CE sequencing may become the newstandard in sequencing based diagnostic technol-ogy. Additional systems are available up to 96capillary systems (AB Prism-3700, SpectrumedixHTS9610 and Amersham Pharmacia Biotech,MegaBACE 1000). More recently AB, which hasbeen a pioneer in CE sequencing with the 310-Prism, has released the 3100-Prism, a 16 capillarysystem. As May, the latest version of the CEQ2000, which uses shorter capillary (37 cm vs 57cm) presents similar performance to the Prism-3100 in term of throughput, for half the cost. ACEQ 2000 XL, allows now, during a workingweek, to run at least five plates of 96 sequencingproducts, ie, 120 full genotyping. This level ofautomation also reduces the cost of the assay andconsequently allows a wider prescription of HIVgenotyping and a more efficient realtime thera-peutic monitoring. In addition to current resis-tance testing recommendations, ie, a case ofprimary infection or multiple therapy failure(Hirsch et al., 2000.), other clinical indicationscould be considered such as treatment-naive pa-

P. Merel1 et al. / Journal of Virological Methods 98 (2001) 9–1616

tients with established infection, first therapy fail-ures and new treatments after structured interrup-tion therapy. Moreover, concomitant viral loadmeasure and HIV genotyping every 3 months maybe now investigated for the follow-up of patientswith virological failure.

There is growing evidence that appropriate useof these tools, in association with drug treatmenthistory, viral load, medication tolerance, futureadherence likelihood and concomitant treatment,can help design more effective drug regimens innaive and treated patients.

Acknowledgements

We are indebted to Jim Osborne, BeckmanCoulter Inc. for his support and technicalassistance.

References

Baxter, J., Mayers, D., Wentworth, D., Neaton, J., Merigan,T. And the CPCRA 046 Study team for the Terry BeirnCommunity programs for clinical research on AIDS(CPCRA), 1999. Final results of CPCRA 046: a pilot studyof antiretroviral management based on plasma genotypicantiretroviral resistance testing (GART) in patients failingantiretroviral therapy. Antiviral Ther. 4 (Suppl. 1), 43.

Collin, G., Descamps, D., Telles, F., Matheron, S., Obry, V.,Costagliola, D., Brun-Vezinet, F., 2000. Differences inprotease and reverse transcriptase sequences between theTruGene HIV-1 genotyping kit (Visible Genetics) and theViroSeq genotyping system (PE Applied Biosystem). An-tiviral Ther. 5 (Suppl.), 53.

Deeks, S., Parkin, N., Petropoulos, C., Grant, R., Volberding,P., Whitcomb, J., Tian, H., Wrin, T., Limoli, K., Drews,B., Warmerdam, M., Hellmann, N., 1998. Correlation ofbaseline phenotypic drug susceptibility with 16 weeks viro-logic response in a pilot combination therapy study inHIV-infected patients who failed indinavir therapy. Antivi-ral Ther. 3 (Suppl. 1), 36–37.

Deeks, S.G., Heilman, N.S., Grant, R.M., Parkin, N.T.,Petropoulos, C.J., Becker, M., Symonds, W., Chesney, M.,Volberding, P.A., 1999. Novel four-drug salvage treatmentregimens after failure of human immunodeficiency virustype 1 protease inhibitor-containing regimen. J. Infect. Dis.179, 1375–1381.

Durant, J., Clevenbergh, P., Halfon, P., Delgiudice, P., Porcin,S., Simonet, P., Montagne, N., Boucher, C.A., Shapiro,J.M., Dellamonica, P., 1999. Drug-resistance genotyping inHIV-1 therapy: the VIRADAPT randomised controlledtrial. Lancet 353, 2195–2199.

Harrigan, P., Montaner, J., Hogg, R., Yip, B., Hertogs, K.,Pauwels, R., Bloor, S., Larder, B., 1998. Baseline resistanceprofile predicts response to SQV/RTV therapy in a com-munity setting. Antiviral Ther. 3 (Suppl. 1), 38.

Hirsch, M.S., Brun-Vezinet, F., D’Aquila, R.T., Hammer,S.M., Johnson, V.A., Kuritzkes, D.R., Loveday, C., Mel-lors, J.W., Clotet, B., Conway, B., Demeter, L.M., Vella,S., Jacobsen, D.M., Richman, D.D., 2000. Antiretroviraldrug resistance testing in adult HIV-1 infection: recom-mendations of an International AIDS Society-USA panel.JAMA 283, 2417–2426.

Hoover, M.L., Wentworth, D.N., Neaton, J.D., Mayers, D.L.,Merigan, T.C., Winters, M.A., Baxter, J.D. and theCPCRA 046 Study team for the Terry Beirn Communityprograms for clinical research on AIDS (CPCRA), 2000. Ablinded comparison of two sequencing assays for resistancemutations in HIV-1 protease and reverse transcriptasegenes in the GART study (CPCRA 046), Antiviral Ther., 5(Suppl.), 60.

Lanier, R., Danehower, S., Daluge, S., Cutrell, A., Tisdale,M., Pearce, G., Spreen, B., Lafon, S., Kemp, S., Bloor, S.,Larder, B., 1998. Genotypic and phenotypic correlates ofresponse to abacavir (ABC, 1592). Antiviral Ther. 3(Suppl. 1), 36.

Lorenzi, P., Opravil, M., Hirschel, B., Chave, J.P., Furrer,H.J., Sax, H., Perneger, L., Perrin, T.V., Kaiser, L., Yerly,S., 1999. Impact of drug resistance mutations on virologicresponse to salvage therapy. AIDS 13, F17–F21.

Schuurman, R., Demeter, L., Reichelderfer, P., Tijnagel, J., deGroot, T., Boucher, C., 1999. Worldwide evaluation ofDNA sequencing approaches for identification of drugresistance mutations in the HIV-1 reverse transcriptase. J.Clin. Microbiol. 37, 2291–2296.

Simpson, K.N., 2000. Budget impact of the integration ofgenotypic antiretroviral resistance testing into HIV treat-ment guides. In: 7th Conference on Retroviruses and Op-portunistic Infections, San Francisco, CA, p. 789 Abstract.

Wilson, J.W., Bean, P., Robins, T., Graziano, F., Persing,D.H., 2000. Comparative evaluation of three human im-munodeficiency virus genotyping systems: the HIV-Geno-typR method, the HIV PRT GeneChip Assay, and theHIV-1 RT Line Probe Assay. J. Clin. Microbiol. 38,3022–3028.

Zolopa, A.R., Schafer, R.W., Warford, A., Montoya, J.G.,Hsu, P., Katzenstein, D., Merigan, T.C., Efron, B., 1999.HIV-1 genotypic resistance patterns predict response tosaquinavir-ritonavir therapy in patients in whom previousprotease inhibitors therapy had failed. Ann. Intern. Med.131, 813–821.