Journal of Medical Virology 75:412–420 (2005)

TRUGENE Sequencing Versus INNO-LiPA forSub-Genotyping of HCV Genotype-4

Abdel Rahman N. Zekri,1* Hanaa M. Alam El-Din,1 Abeer A. Bahnassy,2 Amal M.R. El-Shehabi,3

Heba El-Leethy,1 Ashraf Omar,4 and Hussein M. Khaled5

1Virology and Immunology Unit, Cancer Biology Department, National Cancer Institute, Cairo, Egypt2Tissue Culture Unit, Pathology Department, National Cancer Institute, Cairo, Egypt3Biochemictry Department, Kasr El-Aini School of Medicine, Cairo, Egypt4Tropical Medicine Department, Kasr El-Aini School of Medicine, Cairo, Egypt5Medical Oncology Department, National Cancer Institute, Cairo University, Cairo, Egypt

Hepatitis C virus genotypes and subtypes deter-mination is an important factor forunderstandingtheepidemiologyof thevirus, in thepre-treatmentevaluation of the patients and in defining bettertreatment strategies. In the present study, wecompared two commercially available assaysfor HCV genotyping: the reverse hybridizationbased Innogenetics INNO-LiPA HCV II and thedirect sequencing by TRUGENE assay. The studyincluded 31 HCV-RNA positive Egyptian patients;18 patients with chronic active hepatitis, 8 withHCC, and 5 with cirrhosis. Using the TRUGENEgenotyping test, all the samples had genotype4 (100%) and subtyped as 4a in 18/31(58%), 4c in10/31 (32%), 4e in 1/31 (3%), 4a/c in 1/31 (3%), and4g in 1/31 (3%). Using the INNO-LiPA assay,30 samples had genotype 4 (97%), and 1 samplehad genotype 1e (3%). One sample showedmixed infection with type 4f and type 1. Only sixsamples were subtypable by INNO-LiPA, threewere genotype 4c/d, and the other three were 4f,4e, and 1e. Seven samples gave reactivity inthe INNO-LiPA of lines 5, 6, 16, 17, 18, which areconsidered untypable by the interpretation chartbut considered to be a rare HCV genotype 4 bythe manufacturer. At the genotype level, therewas a 97% concordance between TRUGENEsequencing and INNO-LiPA, but at the subtypelevel the concordance rate was 3% only. Weconclude that the TRUGENE genotyping assay isa reliable test for HCV genotyping for the detec-tion ofmajor types and subtypes detection, whileINNO-LiPA is agood test at the genotype level butunreliable for subtyping especially in the Egyp-tian population. This is mainly due to the highdiversity of genotype 4, which is the most pre-valent genotype in Egypt. J. Med. Virol. 75:412–420, 2005. � 2005 Wiley-Liss, Inc.

KEY WORDS: hepatitis C virus; sequence;Egypt; genotype 4

INTRODUCTION

Hepatitis C virus (HCV) is an enveloped positivesingle-stranded RNA virus that causes chronic hepati-tis and end-stage liver disease worldwide. It is also amajor cause of hepatocellular carcinoma [Sherlock andDooley, 2002]. There is a high incidence of anti-HCVseropositivity in the Egyptian population, with anoverall age-adjusted prevalence of HCV antibodies of21.9% [Frank et al., 2000]. HCV demonstrates a highdegree of sequence variability resulting in the formationof quasispecies. However, the levels of heterogenecitydiffer considerably among the various regions of thevirus, ranging from19% in the 50 untranslated region (50

UTR) to 50% ormore in the E1 region [Chan et al., 1992;Honda et al., 1994; Bukh et al., 1995; Simmonds, 1995].Based on the analysis of variability in the viral genome,many genotypes and subtypes are defined. Themethodsdeveloped to investigate HCV genotypes are based onmolecular biology or serological techniques. The firstincludesdirectnucleicacid sequencing [Simmondsetal.,1993], reverse hybridization line probe assay (LiPA)[Stuyver et al., 1996], subtype-specific reverse tran-scription [Okamoto et al., 1992], DNA restrictionfragment length polymorphism [Nakao et al., 1991],whereas serotyping identifies type specific antibodies

*Correspondence to: Abdel Rahman N. Zekri, M.Sc., Ph.D.,Virology and Immunology Unit, Cancer Biology Department,National Cancer Institute, Fom El-Khalig, Cairo 11796, Egypt.E-mail: ncizakri@starnet.com.eg

Accepted 12 November 2004

DOI 10.1002/jmv.20293

Published online in Wiley InterScience(www.interscience.wiley.com)

� 2005 WILEY-LISS, INC.

against HCV genotype dependant pitopes [Dixit et al.,1995].

The prevalent genotype in Egypt is type 4, with thepresence of other genotypes. Genotype 4 was the com-monest (73%) followed by genotype 1 (26%) whereasmixed infection was found in 15.7% of the studiedsamples [Zekri et al., 2000].

Many studies have used the sequence analysis derivedfrom the 50 UTRofHCV to determine the genotype of thevirus [Chen and Weck, 2002; Anderson et al., 2003;Haushofer et al., 2003]. The two most commonly usedassays are the TRUGENE HCV 50 non-coding (50 NC)genotyping sequencing kit, which is a standard directsequencing kit for genotyping determination, and thereverse hybridization Line Probe Assay (INNO-LiPAHCV II; INNOGENETICS, N.V; Zwijnaarde, Belgium).

To date, there are only few comparison studies ofthese two tests, and most of them are reported fromEurope [Ross et al., 2000; Ansaldi et al., 2001; Halfonet al., 2001;Nolte et al., 2001] andNorthAmerica [Zhenget al., 2003] but none from Egypt where there is a highprevalence of HCV-4.

Therefore, the present study was conducted tocompare HCV genotyping by TRUGENE HCV directsequencing analysis of the 50 NC region and the reversehybridization by INNO-LiPA analysis of the sameregion for samples obtained from different parts of theEgyptian community. The concordance, accuracies, andsensitivities of both assays were assessed in this clinicalevaluation.

PATIENTS AND METHODS

The study included 31 subjects with anti-HCVpositive and high levels of SGPT & SGOT for more than6 months. All cases were subjected to complete historytaking and thorough clinical examination. Liver func-tion tests and liver biopsy were performed. Sampleswere tested for anti-HCVusinga commercially availableEIA.

Detection of HCV

1—RNA extraction. HCVRNAwas extracted fromthe sera by the silica method as previously described[Boom et al., 1990].

2—RT-PCR of HCV. RT and PCR were performedwith a primer pair selected from the highly conserved5-UTR of HCV genome [Choo et al., 1989]. All stepswere done as previously described [Zekri et al., 1995].The following sequenceswere used as antisense primersfor c-DNA synthesis HCV-6 [5-ACC-TCC nucleotides(NT) 319–324]. The internal primers were RB6A andRB6B for amplification of 266 bp of the 5-UTR, RB6A [5-GTG AGG AAC TAC TGT CTT CAC G-3 (NT 47–68)],and RB6B [5-ACT CGC AAG CAC CCT ATC AGG (NT292–312)-3] [Zekri et al., 1995].

All samples were analyzed twice for HCV RNA by theRT-PCR on different days with identical results. Uponcompletion of the amplification reaction, 10 ml of eachPCR reaction product was analyzed by electrophoresis

through a 1.2%agarose gel in Tris-Acetate-EDTAbuffer(pH 8.0) and ethidium bromide staining. DNA wastransferred from the gel onto nitrocellulose filter usingalkaline buffer (4N NaOH). The transferred DNA wascross linkedby incubation for 2–3hr at 808Cand theblotwas then hybridized with an internal probe [Zekri et al.,1995].

HCV Genotyping

All clinical samples were genotyped with both theLiPA and TRUGENE50 NCkits from aliquots of a singleRT-PCR sequence analysis.

1—INNO-LiPA II, genotyping. The line probeassay was used to assess HCV genotypes using kitsprovided by INNOGENETICS, N.V. The 5-UTR regionwas amplified using nested PCR with biotinylatedprimers. The labeled amplicon was allowed to hybridizeand mounted on a strip. After stringent washing, strep-tavidin labeled with alkaline phosphatase was used totrace thehybridizedproducts, andnitroblue tetrazoliumand 5-bromo-4-chloro-3-indoyl-phosphate were usedas a substrate according to the manufacturer’s instruc-tions. The probe reactivity patterns were interpretedusing the chart provided by the manufacturers [Zekriet al., 2000].

2—HCV genotyping by sequencing. The TRU-GENEHCV 50 NC genotyping kit, Visible Genetics, Inc.(Toronto, Ontario, Canada) was used in conjunctionwith the OpenGene DNA sequencing system. All HCVPCR products were run on a 1% agarose gel electro-phoresis. A positive control of known standard sequenceand a negative HCV control provided with the kit wereutilized as controls in each run. The specific band wascut from the gel and purified by silica protocol [Boomet al., 1990], and sequenced by CLIP sequencing whichallows both directions of the target amplicon to besequenced simultaneously in the same tube using twodifferent dye-labeled primers (Cy5.0 andCy5.5) for eachreaction.

This method provides sequence information forboth positive and negative DNA strands from a singlereaction. The forward and reverse sequences are com-bined to form a query sequence. The query sequence isthen compared to previously characterized isolates inthe TRUGENE HCV 50 NC Module of the OpenGenesoftware system in order to determine the HCV geno-type of the sample. Gene Objects software analyzeschromatograms from each sample; the final 50 UTRsequence was obtained from the comparison of bothsequenced strands. This information is compared withdeposited HCV sequences by Gene LibrarianTM soft-ware with a minimal concordance of 98%. The genotypeassignments of these sampleswere confirmedbyBLASTsearches.

Nucleotide sequence accession numbers. Thenucleotide sequences reported here are deposited inThe National Center for Biotechnology Information/National Institute of Health GenBank nucleotide se-quence database (accession no. AY624965–AY624986).

TRUGENE Sequencing Versus INNO-LiPA 413

RESULTS

All of the 31 HCV-RNA positive samples tested bythe TrueGene sequencing were genotype 4 (100%).They were subtyped as 4a (65%), 4c (29%), 4a/c (3%),and 4g (3%) (Table I).

Figure 1 shows 50 NC region sequence analysis andFig. 2 showed the phylogenetic analysis of all studiedsamples in conjunction with three known GenBanknucleotide sequences with Blast database. Comparisonbetween the sequences obtained in the present studyand all previously reported 50 UTR sequences revealed anew 50 UTR mutation within HCV genotype-4 andconfirm the heterogeneity of this type in the Egyptianpopulation. On the other hand, most of these mutationswere randomly distributed, however; they still main-tained the postulated secondary structure of the 50 UTR

TABLE I. Comparison of HCV Genotyping ResultObtained With the LiPA and TRUGENE 50 NC Kit for

31 Clinical Specimens

LiPA type

TRUGENE sequencing

4a 4c 4e 4g 4a/c Total

4 12 4 1 1 184c/d 2 1 34e 1 11e 1 14fþ 1 1 1NTa 2 5 7Total 18 10 1 1 1 31

aNT, not typeable by LiPA.

Fig. 1. The alignment of all cases typed by both INNO-LiPA and TRGUENE sequence analysis withmultiple sequence alignment of HCV-genotype-4 from the gene data bank (X78866, X78865, AF142056)using CLUSTALW analysis.

414 Zekri et al.

genomic RNA and revealed the strong conservation ofthe genotypes.

By the INNO-LiPA, the samples were found to begenotype 4 in 23 (74%) samples and genotype 1 in 1 (3%)sample. Seven samples (23%) gave reactivity of lines 5,6, 16, 17, 18 which are considered untypable by theinterpretation chart. However, by contacting the man-ufacturer, they considered them to be rare genotype 4patterns, which will most probably be added to thefuture update of the interpretation chart. The geno-type assignments of these samples were confirmed bythe blast search. The concordance between TrueGenesequencing and INNO-LiPA was 74% (23/31 samples)at the main genotype level (Fig. 1A). By adding theseven untypable samples, which were considered by themanufacturer as type 4, the concordance rises to (97%).The cladograms of these sequences with known type 4sequences from the GenBank are shown in Figure 1B.

Subtypingwas possible in six cases only by the INNO-LiPA assay, of which three (9.7%) were 4c/d, one (3%) 4f,

one (3%) 4e, and one (3%) 1e (Table I). Figure 1C showsthe cladogram of five of these caseswith type 4. Only onecase showed mixed infection by INNO-LiPA, however itshowed genotype-4a by TrueGene.

One sample only (AY624963) showed discordancebetween the two tests at the genotype level. This samplewas classified as genotype 1e by INNO-LiPA and asgenotype 4a by TRUGENE. The cladogram of thissequencealongwith type1 sequences fromtheGenBankis shown in Figure 1D.

The demographic, clinical, and genotype features ofthe patients in relation to liver biopsy data are shownin Table II. The incidence of ascites, jaundice, andsplenomegaly was significantly higher in HCC thanin CAH patients (P¼ 0.000, 0.02, 0.03, respectively).HCC patients were mainly from Metropolitan CairoArea (P¼0.03), whereas those with CAH were mainlyfrom Delta area (P¼<0.001).

Table II also shows the comparison between genotype4a and 4c in relation to different study parameters.

Fig. 1. (Continued )

TRUGENE Sequencing Versus INNO-LiPA 415

No significant abnormalities were detected in relationto the two subtypes.

DISCUSSION

The identification of different HCV genotypes, sub-types, and isolates is helpful in understanding theevolution and the epidemiology of the virus in relationto patient age, area, risk factors, and degree of liverdisease [Icardi et al., 1997]. Although sequence analysisis considered the ‘‘gold standard’’ for HCV genotypedetermination, it is expensive, time consuming, andinconvenient for routine use. Analysis of the 50 UTR bythe INNO-LiPAHCV II provides a fast and easymethodfor determination of theHCVgenotype [ChenandWeck,2002]. To date, there are only few comparison studies ofthese two tests and most of them are reported fromEurope [Ross et al., 2000; Ansaldi et al., 2001; Halfonet al., 2001;Nolte et al., 2001] andNorthAmerica [Zheng

et al., 2003]. Our study is the first comparison betweenthe two methods using clinical samples from Egypt.

The results of the present study demonstrate a goodconcordance between the two genotyping methods atthe genotype level, but the possibility to analyze thesequence allows an absolute precision in the classifica-tion of the subtypes. Sequencing analysis by TRUGENEmethod allowed subtyping of all samples, whereas onlyfive samples could be subtyped by the INNO-LiPA test.We found a very high level of agreement (97%) betweenTrueGene sequencing system and INNO-LiPA at thegenotype level. Other studies showed similar results[Chen andWeck, 2002;Gargiulo et al., 2003;Nolte et al.,2003; Zheng et al., 2003]. At the subtype level there wasonly one sample (3%) that showed concordance betweenthe two methods. This disagreement was reported byothers [Chen and Weck, 2002; Nolte et al., 2003]. Thelack of agreement reported in this study at the subtypelevel is not surprising considering the high diversity of

Fig. 1. (Continued )

416 Zekri et al.

genotype 4, especially in the Egyptian population aspreviously shown by others [Zekri et al., 2001] compar-ing sero- and geno-typing methods, The detection of alarge number of individual mutated strains by hybridi-zation assay such as INNO-LiPA in a single test isdifficult since it requires very high stringent conditions[Ansaldi et al., 2001]. Also the lack of agreementbetween the subtypes is not unusual considering thatbothmethods target the50NCregion.The50NCregion isamong the most conserved regions of the viral genome,and in several cases, only one or two nucleotide changesdistinguish unique subtypes [Smith et al., 1995]. It hasbeen suggested that clinical laboratories should not callHCV subtypes from analysis of the 50 NC regardless ofthe method employed because of the inherent inaccu-racy of the calls [Nolte et al., 2003]. Also they stated thatthere are no recognized subtype-specific differences indisease progression or response to therapy that wouldwarrant these designations.

In our study, subtypes 4a and 4c were the mostprevalent; this was similar to that previously reported

[Zekri et al., 2000]. We could not correlate between thedifferent subtypes and clinical picture, gender, resi-dence, or liver biopsy results due to the low number ofcases studied. In Egypt, the response of genotype 4to interferon treatment was reported in many studies[El-Zayadi et al., 1999; Esmat et al., 2002; Thakeb et al.,2003], however there are no available data of theresponse of subtypes 4a and 4c to interferon treatment.

Regarding the patient with cirrhosis that showednon-concordant results, it was classified as type 1e byINNO-LiPA and genotype 4a by the Visible Geneticssequencing system. A similar result was previouslyobtained [Ansaldi et al., 2001]. The mistyping could beascribed to the close phylogenetical similarity betweengenotype 1 and 4, since the difference in the regionanalyzed is only six nucleotides [Ansaldi et al., 2001].

From this study, we concluded that TRUGENEsequencing for HCV genotyping and INNO-LiPA pro-vide reliable results at the type level and have almostsimilar analytical sensitivities regarding genotype 4.TRUGENE method cannot, however, detect mixed

Fig. 1. (Continued )

TRUGENE Sequencing Versus INNO-LiPA 417

Fig. 2. The first cladogram shows all cases (A); the second shows the seven INNO-LiPA untypable (B);the third shows the five genotype 4 subtypable by INNO-LiPA (C); and the fourth (D) shows the case withgenotype 1 with four other HCV genotype 1 from the gene data bank (Z36518–Z36520).

418 Zekri et al.

infection as the INNO-LiPA method. Although, theTRUGENE 50 NC method is more technically complexto perform than LiPA, it is more detailed informativemethod in the level of the subtype especially inEgyptianpopulations with a very high heterogeneity of thesutype-4. It is also the best method for the epidemiolo-gical investigations. In addition, TRUGENE 50 NCsystem can be easily updated, as new HCV sequenceinformation becomes available of the HCV genome. Astreatment of HCV is dependent on main genotype andnot subtype, INNO-LiPA test can be used for geno-typing of HCV until more studies can detect significantdifference in response to interferon in different subtypesin genotype 4, the most prevalent in Egypt. Futureanalysis is needed in order to correlate subtyping withTRUGENE with response to interferon treatment inEgyptian patients with type 4, since all previouslypublished data in this field was done by INNO-LiPAsequencing method.

REFERENCES

Anderson J, Simonetti J, Fisher D, Williams J, Yamamura Y,Rodriguez N, Sullivan DG, Gretch DR, McMahon B, Williams KJ.2003. Comparison of different HCV viral load and genotypingassays. J Clin Virol 28:27–37.

Ansaldi F, Torre F, Bruzzone B, Piciotto A, Crovari P, Icardi G. 2001.Evaluation of new hepatitis C virus sequencing assay as a routinemethod for genotyping. J Med Virol 63:17–21.

Boom R, Sol CJA, Salimans MMM, Jausen CL, Wertheim-Van DillenPME, Van der Noordaa J. 1990. Rapid and simple method forpurification of nucleic acids. J Clin Microbiol 28:495–503.

Bukh J,Miller R, Purcell R. 1995. Genetic heterogenicity of hepatitis Cvirus: Quasispecies and genotypes. Semin Liver Dis 15:41–63.

Chan SW, McOmish F, Holmes EC, Dow B, Peutherer JF, Follett E,Yap PL, Simmonds P. 1992. Analysis of a newhepatitis C virus typeand its phylogenetic relationship to existing variants. J Gen Virol73:1131–1141.

Chen Z, Weck K. 2002. Hepatitis C virus genotyping: Interrogation ofthe 50 untranslated region cannot accurately distinguish genotypes1a and 1b. J Clin Microbiol 40:3127–3134.

Choo QL, Kuo G, Weiner AJ, Overby LR, Bradley DW, Houghton M.1989. Isolation of cDNA clone derived from a blood-borne non-A,non-B viral hepatitis genome. Science 244:359–362.

Dixit V, Quan S, Martin P, Larson D, Brezina M, DiNello R, Sra K,Lau JY, Chien D, Kolberg J, et al. 1995. Evaluation of a novelserotyping system for hepatitis C virus: Strong correlation withstandard genotyping methodologies. J Clin Microbiol 33:2978–2983.

El-Zayadi A, Selim O, Haddad S, Simmonds P, Hamdy H, Badran HM,Shawky S. 1999. Combination treatment of interferon alpha-2band ribavirin in comparison to interferon monotherapy in treat-ment of chronic hepatitisCgenotype 4patients. Ital JGastroenterolHepatol 31:472–475.

Esmat G, Abuzied A, Abdel-Aziz F, et al. 2002. Treatment with peg-interferon alfa 2b (peg-IFN) plus ribavirin compared to interferonalfa 2b (IFN alfa-2b) plus ribavirin on subjects with chronichepatitis C infected with HCV genotype 4. Hepatology , AASLDAbstr 364A.

Frank C, Mohamed MK, Strickland GT, Lavanchy D, Arthur RR,Magder LS, El Khoby T, Abdel-Wahab Y, Aly Ohn ES, Anwar W,Sallam I. 2000. The role of parenteral anti-schistosomaltherapy in the spread of hepatitic C virus in Egypt. Lancet 355:887–891.

Gargiulo F, De FrancescoM, Pinsi G, Pallara C, Terlenghi L, PerandinF, Manca N. 2003. Determination of HCV genotype by directsequence analysis of quantitative PCR products. J Med Virol69:202–206.

HalfonP, Trimoulet P,BourliereM,KhiriH, deLedinghenV,CouzigouP, Feryn JM, Alcaraz P, Renou C, Fleuy HJA, Ouzan D. 2001.Hepatitis C virus genotyping based on 50 noncoding sequenceanalysis (Trugene). J Clin Microbil 39:1771–1773.

Haushofer A, Berg J, Hauer R, Trubert-Exinger D, Stekel HG, KesslerHH. 2003. Genotyping of hepatitis C virus. Comparison of threeassays. J Clin Virol 27:276–285.

Honda M, Kaneko S, Sakai A, Unoura M, Murakami S, Kobayashi K.1994. Degree of diversity of hepatitis C virus quasispecies andprogression of liver disease. Hepatology 20:1144–1150.

Icardi G, Bonanni P, Wicks R, Faraldi L, Roccatagliata A, Santo M,Orione L, De Conca V, Crovari P. 1997. Characterization ofhepatitis C virus genotypes in Italian subjects at different riskinfection. In: Rizzetto M, Purcell H, Gerin JL, Verme G, editors.Viral hepatitis and liver disease. Torino:EdizioneMinerva. pp598–600.

Nakao T, Enomoto N, Takada N, Takada A, Date T. 1991. Typing ofhepatitis C virus genomes by restriction fragment length poly-morphism. J Gen Virol 72:2105–2112.

Nolte RS, Green A, Fiebelkon K, Caliendo A. 2001. Clinical evaluationof two methods for genotyping hepatitis C virus based on50noncoding (NC) region sequence analysis. J Mol Diag 3:204.

Nolte FS, Green AM, Fiebelkorn KR, Caliendo AM, Sturchio C,Grunwald A, Healy M. 2003. Clinical evaluation of two methodsfor genotyping hepatitis C virus based on analysis of the 50 non-coding region. J Clin Microbiol 41:1558–1564.

TABLE II. Demographic, Clinical And Genotype Features of the Study Groups

Parameter

Diagnosis

P-value

Genotype

P-valueCAH

(n¼ 18)Cirrhosis(n¼ 5) HCC (n¼ 8) 4a (No. %) 4c (No. %)

SexM/F 18/0 4/1 5/3 0.01 18/2 8/1 0.928Age (mean�SD) 47� 10 56� 10 53� 12 0.08 51.5� 11.7 46.3� 8.7 0.22

Liver stateHepatomegaly 14 (78%) 3 (60%) 8 (100%) 0.12 17 (71%) 7 (29%) 0.634Ascites 1 (5.5%) 2 (40%) 8 (100%) 0.000 8 (72.7%) 3 (27.3%) 0.732Jaundice 2 (11%) 1 (20%) 5 (62.5%) 0.02 5 (62.5%) 3 (37.5%) 0.642L.L. oedema 1 (5.5%) 2 (40%) 3 (37.5%) 0.12 3 (75%) 1 (25%) 0.779Splenomegaly 8 (44%) 5 (100%) 8 (100%) 0.03 15 (79%) 4 (21%) 0.109

ResidenceMCA 3 (16.7%) 1 (20%) 6 (75%) 6 (60%) 4 (40%)Delta 14 (78%) 2 (40%) 2 (25%) 12 (71%) 5 (29%)Others 1 (5.5%) 2 (40%) 0 (0%) 2 (100%) 0 (0%)P-value <0.001 0.95 0.03 0.523

CAH, chronic active hepatitis; HCC, hepatocellular carcinoma; L.L. oedema, lower limb oedema; MCA, Metropolitan Cairo Area.P-value <0.05 is significant.

TRUGENE Sequencing Versus INNO-LiPA 419

Okamoto H, Sugiyama Y, Okada S, Kurai K, Akahane Y, Sugai Y,Tanaka T, Sato K, Tsuda F, Miyakawa Y, et al. 1992. Typinghepatitis C virus by polymerase chain reaction with type-specificprimers: Application to clinical surveys and tracing infectioussources. J Gen Virol 73:673–679.

Ross RS, Viazov SO, Holtzer CD, Beyou A, Monnet A, Mazure C,Roggendorf M. 2000. Genotyping of hepatitis C virus isolates usingCLIP sequencing. J Clin Microbiol 38:3581–3814.

Sherlock S, Dooley J. 2002. Diseases of liver and biliary system, 11thedn. Oxford UK: Blackwell scientific publication. Chapter (18).pp 305–320.

SimmondsP. 1995.Variability of hepatitisCvirus.Hepatology21:570–583.

Simmonds P, Holmes EC, Cha TA, Chan SW, McOmish F, Irvine B,Beall E, Yap PL, Kolberg J, Urdea MS. 1993. Classification ofhepatitis C virus into six major genotypes and a series of subtypesby phylogenetic analysis of the NS-5 region. J Gen Virol 74:2391–2399.

Smith DB, Mellor J, Jarvis LM, Davidson F, Kolberg J, Urdea M,Yap PL, Simmonds P. 1995. Variation of the hepatitis C virus 50

noncoding region: Implications for secondary structure, virusdetection, and typing. J Gen Virol 76:1749–1761.

Stuyver L,Wyseur A, vanArnhemW,Hernandez F,Maertens G. 1996.Second-generation line probe assay for hepatitis C virus geno-typing. J Clin Microbiol 34:2259–2266.

Thakeb F, Omar M, El-Awady M, et al. 2003. Randomized controlledtrial of peg-interferon alfa-2a plus ribavirin for chronic hepatitisC virus genotype 4 among Egyptian patients. Hepatology38(Suppl 1):252 (Abstract).

Zekri A, Bahnassay A, Khaled H, Mnsour O, Attia MA. 1995.Comparative analysis of different PCR techniques for detectionof HCV in hepatocellular carcinoma patients. Cancer J 8:331–335.

Zekri AR, Bahnassy AA, Shaarawy SM, Mansour OA, Maduar MA,Khaled HM, El-Ahmadi O. 2000. Hepatitis C virus genotyping inrelation to neu-oncoprotein overexpression and the development ofHepatocellular carcinoma. J Med Microbiol 49:89–95.

Zekri A, Bahnassy A, Ramadan A, El-Bassuoni A, Badran A, MadwarMA. 2001. Hepatitis C virus genotyping versus serotyping inEgyptian patients. Infection 29:24–26.

Zheng X, Pang M, Chan A, Roberto A, Warner D, Yen-Liberman B.2003. Direct comparison of hepatitis C virus genotypes tested byINNO-LiPA HCV II and TRUGENE HCV genotyping methods.J Clin Virol 28:214–216.

420 Zekri et al.