Molecular techniques for clinical diagnosticvirology

Steven J Read, David Burnett, Colin G Fink

A decade after the first studies were performedit is justifiable to claim that molecular tech-niques have revolutionised the work of theclinical virology laboratory. Hitherto, the roleof the virology laboratory was often a retro-spective diagnosis based on virus isolation andserology. Nevertheless, the epidemiologicaldata collected in this way justified the contin-ued activity of clinical virology. The recentmolecular revolution in laboratory methodshas been timely because it has been in parallelwith the emergence of new pathogens that havepresented the clinical virologist with freshdiagnostic challenges. A concurrent develop-ment of specific antiviral compounds hasincreased the potential of rapid laboratoryinvestigation to contribute to the managementof acutely ill or immunosuppressed patients.

Molecular assays for the detection of micro-organisms can be designed even when onlypartial nucleic acid sequence information isavailable. This is valuable when identifying anddiagnosing new diseases and emerging patho-gens because there is the possibility of a rapiddevelopment of assays in house. However, thestandardisation of assay design is a continuingproblem because of the large number of nucleicacid sequences that can be used as a target forthe detection of any one virus. Many of themethodologies associated with molecular tech-niques have now been incorporated intocommercial kits, but in house molecular assayscontinue to be used. In house assays are oftenwell evaluated and perform to a high standard,but diVering assay sensitivities and specificitieshave been highlighted by quality assessmentprogrammes. Some of the variations in per-formance can be accounted for by assay design,although it is probable that some in houseassays do not travel well and perform badlybecause laboratory staV are inexperienced withmolecular techniques. Molecular techniques intheir present format, even in commercial kits,require a considerable degree of operator skill.If a wider use of molecular techniques in a dis-trict hospital setting is considered desirable,more robust protocols that consider theextreme sensitivity of amplification assays,together with more standardisation of method-ology, will be required. Most in house assaysare unsuited to high throughput testing and aredemanding of staV time; these problems arenow being dealt with by developments in assayautomation.

MethodologiesTARGET AMPLIFICATION TECHNIQUES

The most sensitive molecular techniques usean enzymatic step to amplify the target nucleicacid before detection of the specific sequence.

Clinical material may be rich in humangenomic DNA, but it is possible to amplify onlya very few target viral genomes present tobecome the dominant, easily detected se-quence. A variety of target amplificationtechniques has been developed but those mostwidely used and commercially developed arethe polymerase chain reaction (PCR), ligasechain reaction (LCR), and nucleic acid se-quence based amplification (NASBA).

PCRPCR was the first nucleic acid amplificationtechnique to be described and, perhaps be-cause of its widespread use in research applica-tions, has become the most widely usedmolecular diagnostic technique in clinicalvirology. There are several reviews, chapters,and books devoted to its methodology andapplication.1 2 The exponential amplification ofa single target sequence is theoretically possibleso that exquisitely sensitive detection isachieved. PCR has been made more versatileby the development of several variants of thebasic method; RNA genomes can be detected iffirst converted into a complementary DNAcopy; several nucleic acid sequences can bedetected simultaneously using a cocktail ofprimers; the technique can be made more sen-sitive and specific by using double amplifica-tion with appropriately designed “nested”primers, or the amplification may be made lessspecific to detect divergent genomes or par-tially characterised sequences.

Alternative target amplification techniquesThe development of commercial assays hastaken place only for relatively high throughputapplications, such as the detection of Chlamy-dia trachomatis and human immunodeficiencyvirus (HIV). PCR is covered by patent rightsand so the major commercial development ofthis technique has been restricted to onemanufacturer. Neither of the alternative targetamplification techniques is widely used as thebasis for in house assays, but they have beendeveloped for use in commercial kits because ofthe restrictions on PCR use. LCR, developedby Abbott Laboratories, uses a thermostableligase to join two oligonucleotides to each otherwhen they are hybridised to their respectivetarget sequences. Amplification of nucleic acidsequences by LCR can be exponential becauseeach round of synthesis can double the numberof target molecules. NASBA, developed byOrganon Teknika, has the advantage of beingan isothermal process, so that thermal cyclingequipment is not required. Compared with theother nucleic acid amplification methods, it isparticularly suited to the detection of RNA

J Clin Pathol 2000;53:502–506502

Micropathology Ltd,University of WarwickScience Park, BarclaysVenture Centre, SirWilliam Lyons Road,Coventry CV4 7EZ, UKS J ReadD BurnettC G Fink

Correspondence to:Dr Reademail:micropathology@telinco.co.uk

Accepted for publication10 November 1999

on 29 July 2018 by guest. Protected by copyright.

http://jcp.bmj.com

/J C

lin Pathol: first published as 10.1136/jcp.53.7.502 on 1 July 2000. D

ownloaded from

viruses because an RNA polymerase is used toamplify RNA without conversion to comple-mentary DNA.

Signal amplification techniquesThese techniques do not replicate the nucleicacid detected in the assay but rather use ampli-fication of a signal generated by detection ofthe target sequence. Branched DNA (bDNA)technology incorporated into commercial as-says by Chiron Corporation uses multipleDNA probes, first for detection, and subse-quently for increasing the number of potentialbinding sites from which the reporter signal iseventually generated. This approach is lesssensitive than target amplification but has sev-eral advantages. The assay format is well suitedto routine high throughput testing and it is notprone to false positives caused by crosscontamination. It is therefore more readilysuited for use in virology laboratories wherestaV are inexperienced in molecular tech-niques. Importantly, bDNA assays are capableof detecting a broader range of viral genotypesthan PCR based assays, which with their morelimited number of target sequences have arestricted specificity: potentially a problemwhen detecting viral RNA genomes, which canshow high sequence heterogeneity. The bDNAbased Quantiplex assays for the measurementof HIV and hepatitis C virus (HCV) viraemiameasure more accurately a broader spectrumof viral genotypes than the first generationPCR based Amplicor assays.3

Applications of molecular techniquesNEW APPROACHES TO THE DISCOVERY OF VIRUSES

AND DISEASE AETIOLOGIES

Molecular techniques have been instrumentalin the recent discoveries of viruses associatedwith hepatitis.4 The most important of thesecame in 1989 with the partial characterisationof the infectious agent associated with mostcases of non-A non-B hepatitis.5 6 This condi-tion usually presented as post-transfusionjaundice and was long suspected to have a viralaetiology. Nucleic acid from infectious chim-panzee blood was used to produce suYcientquantities of virus encoded proteins for thedevelopment of a diagnostic antibody assay.This molecular approach was successful indeveloping a laboratory diagnostic assay beforethe virus had been isolated and characterised inthe conventional sense; now named hepatitis Cvirus, it has been characterised extensivelyusing molecular techniques. Another virus hasbeen similarly discovered and characterised,7

and although termed hepatitis G virus, a rolefor this virus or group of viruses in human dis-ease remains unclear.8

A powerful PCR based technique known asrepresentational diVerence analysis has beenused to discover a virus associated with cancerin HIV infected individuals.9 Apart fromepidemiological data suggesting an infectiouscause of Kaposi’s sarcoma, there had been nocharacterisation of the newly discovered mem-ber of the herpesvirus group. Human herpes-virus 8, or KS virus, has been detected in tissue

from HIV infected individuals, including thosewith Kaposi’s sarcoma lesions,10 11 and alsofrom sarcoid.12

DIAGNOSTIC PROBLEMS

Traditionally, laboratory diagnostic techniquesin virology have relied on an ability topropagate infectious virus from clinical mate-rial in cell culture, or on the detection of a spe-cific antibody. Where virus propagation is notsuccessful, laboratory diagnosis has remained aproblem because direct detection methods forviral antigen, nucleic acid, or morphology bymicroscopy in clinical material have low sensi-tivity. These diYculties, together with the timeinterval required to detect a specific immuneresponse and the interpretation of antibodyvalues for viruses that cause latent infections,has caused continuing diagnostic dilemmas forthe clinical virologist. A reliable diagnosis ofHIV and HCV infection is usually achieved bythe detection of antibody; however, theseviruses can be transmitted vertically andpassive acquisition of maternal antibody con-fuses confirmation of infection in the neonate.The serological diagnosis of HIV and HCVinfection is also impossible in the “windowperiod” before seroconversion, and this poses aconcern for the safety of blood products: anestimated risk of HCV transmission in westernGermany is 1/20 000 for first time donors and1/200 000 for repeat donors.13

Molecular assays in some circumstancesmight be no more sensitive or specific than tra-ditional techniques, although they may bemore suited to large scale screening pro-grammes. The detection of C trachomatis infec-tion, an “atypical” organism that has tradition-ally been the interest of the virologist, hasevolved from isolation in cell lines, which istechnically diYcult but has good specificity, toantigen detection by enzyme immunoassay orfluorescence antibody test, and now to molecu-lar amplification of pathogen DNA. The com-mercial molecular assays that have been devel-oped are more sensitive than antigendetection,14 and much better suited to routineuse than cell culture isolation. The use ofmolecular detection has led to a growingawareness of the extent of chlamydial infectionand the potential value of a comprehensivescreening programme.

CENTRAL NERVOUS SYSTEM INFECTION

Central nervous system (CNS) viral infections,although recognised clinically, are a diagnosticdiYculty because of the generally low sensitiv-ity of virus isolation methods from cerebrospi-nal fluid (CSF). Herpes simplex virus type 1(HSV-1) is rarely recoverable from the CSF ofpatients with herpes simplex virus (HSV)encephalitis, probably because relatively fewviable virus particles are released into the CSF.For a while, techniques based on the detectionof virus in brain biopsy were attempted to pro-vide a reliable, prompt diagnosis in this illness,but the advent of eVective antivirals of low tox-icity has made invasive techniques diYcult tojustify. In cases of enterovirus meningitis, virusmight be recovered from CSF, although not all

Molecular techniques for clinical diagnostic virology 503

on 29 July 2018 by guest. Protected by copyright.

http://jcp.bmj.com

/J C

lin Pathol: first published as 10.1136/jcp.53.7.502 on 1 July 2000. D

ownloaded from

serotypes grow in laboratory cell lines, andgrowth is generally too slow to aid diVerentiat-ing between viral and bacterial meningitis.Hitherto, patients with viral CNS infectionhave often received only a presumptive diagno-sis based on clinical presentation: without sup-portive laboratory evidence an understandingof the aetiology and epidemiology of viral CNSdisease has been limited.

Molecular amplification techniques havebeen able to overcome many of these problemsthrough the ability to detect just a few copies ofviral nucleic acid directly in CSF. The use ofthe PCR technique for the diagnosis of HSVencephalitis was one of its first applications inthe clinical virology laboratory, and when it wasdemonstrated that PCR could potentially pro-vide sensitive and specific diagnosis,15 the testwas evaluated extensively in several prospectivestudies.16–18 Assays for the other herpesvirusesand enteroviruses have also been developedand found to contribute to laboratory diagnosisand patient management.19–22 The molecularamplification approach to the diagnosis of HSVencephalitis is now recognised as the method ofchoice for this disease,17 23 and also for viralCNS infection in general.24 PCR analysis ofCSF has demonstrated that symptoms of men-ingitis might be caused by the reactivation of

varicella zoster virus (VZV),25 26 or herpes sim-plex virus type 2 (HSV-2)26 27 infection withoutconcurrent skin lesions. HSV-2 infection hasbeen shown to be the cause of Mollaret’smeningitis28 and an important cause of appar-ent sporadic aseptic meningitis; in one study,this virus was the most common aetiologyfound in women with this clinical diagnosis.29

Where traditional diagnostic methods areinsensitive, our understanding of the naturalhistory of viral disease continues to beimproved through the use of molecular tech-niques, and in CNS infection our knowledge ofthe range of viral aetiologies has improvedalongside the potential for early therapeuticintervention.

DIAGNOSIS IN IMMUNOSUPPRESSED PATIENTS

The development of antiviral drugs has high-lighted the potential value to treatment of arapid diagnosis of virus reactivation in immuno-suppressed patients. Cytomegalovirus (CMV)is a major cause of post-transplant morbidityand, because it replicates only slowly inconventional cell culture, methods such asimmunostaining for CMV early antigen ex-pression have been used to speed the detectionof virus replication in inoculated cell monolay-ers in the laboratory. Methods for detectingdirectly either virus specific antigen30 or nucleicacid31 in clinical material have removed theneed for virus cell culture. An early concernthat the high sensitivity of molecular amplifica-tion techniques would result in the detection oflatent herpesvirus genome, so lowering theclinical specificity of virus nucleic acid detec-tion, have been unfounded: the detection ofviraemia has been found to correlate well withactive CMV disease32 33 and CMV PCR is nowoften incorporated into routine post-transplantmicrobiological surveillance of susceptible pa-tients.

VIRAL GENOTYPING

The problem of drug resistant virus pheno-types is becoming more recognised with anincrease in the number of immunosuppressedpatients receiving long term antiviral treat-ment. The use of prophylactic antiviral treat-ment for reactive herpesvirus infection, and theuse of combinations of drugs to suppress viralreplication in HIV infection, has selected formutations conferring drug resistance. Underselective pressure, a drug resistant strain canquickly become the dominant phenotype, butmolecular based assays have improved theclinical management of these patients by facili-tating the detection of mutations conferringresistance. For the HIV genome, a commercialassay is available based on molecular amplifica-tion of the viral genome followed by hybridisa-tion with mutation specific DNA probes. Atpresent, these methods are used for the detec-tion of resistance to the reverse transcriptaseinhibitor class of antiviral compounds, but as arapid and inexpensive approach it is likely to beused for other phenotypic determinants. Asimilar technique has been used for genotypingHCV infection for epidemiological study andfor predicting the response to treatment.34 35

Figure 1 Measurement of hepatitis B virus (HBV) genome in serum using real timemonitoring of polymerase chain reaction (PCR) amplification kinetics of a set of standardson the LightCycler instrument. The PCR product is detected by a double stranded DNAspecific fluorescent dye. Each line represents a PCR amplification. This experimentdemonstrates the amplification of HBV DNA from a 1 log10 dilution series of the HBVDNA positive serum sample standard with an initial concentration of 107 genomes/ml. Theconcentration of HBV genome is proportionate to the PCR cycle at which the fluorescentsignal increases above the background level (represented by the horizontal line).

100.000

10.000

0.100

1.000

0.010361814 16128 10

Cycle number

Flu

ore

scen

ce (

F1)

640 2 282422 26 32 343020

Figure 2 Additional information about the reaction detailed in fig 1 can be obtained fromthe melting curve analysis. Analysis of the melting characteristics of the polymerase chainreaction (PCR) products demonstrates two distinct products: primer dimer, melting at 74°C,and amplified hepatitis B virus (HBV) product melting at 84°C. Primer dimer formationoccurs only at lowest input copy numbers.

13.012.0

10.09.08.07.06.05.04.03.02.01.00.0

–1.0–2.0–3.0

11.0

–4.089.0

80.078.079.077.075.0

76.0

Temperature (°C)

Flu

ore

scen

ce –

d(F

1)/d

T

74.073.065.0

70.071.0

72.068.067.0

66.069.0 85.083.0

82.0 84.087.0

88.086.081.0

504 Read, Burnett, Fink

on 29 July 2018 by guest. Protected by copyright.

http://jcp.bmj.com

/J C

lin Pathol: first published as 10.1136/jcp.53.7.502 on 1 July 2000. D

ownloaded from

VIRAL LOAD

Molecular assays have had an important role inthe study of the natural history of HIVinfection, and subsequently in the developmentof eVective treatment strategies. Quantitativemolecular assays have shown that HIV-1 infec-tion is characterised by high rates of virus pro-duction and clearance of both infected cellsand cell free virions.36 This discovery led to thedevelopment of highly active antiretroviraltreatment, the aim of which is the durablemaximum suppression of virus replication.37

This advance in HIV treatment has been facili-tated by the recent commercial development ofviral load assays. The assessment of viraemia isnow used as a prognostic indicator in HIV,38

HCV,39 and hepatitis B virus40 infections, andis also used in monitoring the eYcacy of anti-viral treatment in systemic viral infections ofimmunosuppressed patients.33

Further developmentsThe introduction of molecular techniques toroutine diagnostic virology laboratories mayhave been held back by the lack of assaysdeveloped commercially, the high initial costs,and the specialist training involved in establish-ing these techniques. Using the technologycurrently available, the areas described mightrepresent their limits in clinical virology;however, it is probable that formats currentlyunder development and further genome se-quencing will spread molecular detection morewidely throughout diagnostic microbiology.

AUTOMATED ASSAY FORMATS

Automation in virology laboratories for serol-ogy assays has been slow to transfer to molecu-lar amplification techniques, but sequencedetection and other downstream processeshave now been automated successfully with theintroduction of colorimetric based detectionmethods. However, if molecular assays are tobecome the methods of choice, there is arequirement for automated procedures forspecimen preparation and nucleic acid extrac-tion, which currently are “hands on” and timeconsuming. An important proposed develop-ment will be the ability to perform molecularassays on an automated platform currentlywidely used for serology assays. If theseprocesses are automated successfully, it seemslikely that for high throughput applicationsmolecular assays will replace many traditionalmicrobiological methods, which will become incontrast, both slow and expensive.

REAL TIME DETECTION AND CHARACTERISATION

OF DNA SEQUENCES WITH FLUORESCENT DYES

An exciting new development is the combina-tion of thermal cycling for PCR nucleic acidamplification and fluorimetry in a singlemachine, such as the LightCycler instrumentfrom Roche Molecular Biochemicals, able tosupport a number of fluorescent chemistries. Itis possible by measurement of the signal fromDNA sequence specific fluorescent dyes tomonitor amplification kinetics in real time,allowing for accurate, rapid, and cheaper viralload tests (figs 1 and 2). These assays have thepotential for accurate measurement over a widerange of viral concentrations. This instrumentdrastically reduces the problem of contamina-tion in PCR because the reaction is analysed ina sealed system without exposure to theenvironment. Viral genotyping and mutationanalysis can also be performed within thesystem by characterisation of the PCR product.The technology facilitates assay automationbecause the amplification, data collection, andanalysis can be performed by a single machineconnected to a computer.

SEQUENCE DETECTION AND CHARACTERISATION

USING OLIGONUCLEOTIDE MICROARRAYS

Microarray “DNA chips” containing immobi-lised oligonucleotide probes can be used toscreen rapidly for a number of target moleculesamplified from clinical material (fig 3). Geneticscreening has encouraged a rapid developmentof this technology,41 42 and microbiologicalapplication has recently been described inscreening for drug resistance in tuberculosis.43

This technology might lead to the developmentof diagnostic kits for specific syndromes—forexample, CNS infection, fever in immunosup-pressed transplant recipients, genitourinary orupper respiratory tract infections. The use ofthis “syndromic” approach for the routinediagnostic laboratory would see a breakdownin the speciality barriers within microbiologylaboratories because the techniques used forthe detection of viruses, bacteria, fungi, anddrug resistant genotypes all become based onthe detection and characterisation of pathogen

Figure 3 A “laboratory on a chip”. A DNA chip containing immobilised arrays ofoligonucleotides for the study of gene expression. Specialised equipment is required forinterpreting the patterns generated by hybridisation of DNA to the array but the techniqueis suited to automated, high throughput applications.

Molecular techniques for clinical diagnostic virology 505

on 29 July 2018 by guest. Protected by copyright.

http://jcp.bmj.com

/J C

lin Pathol: first published as 10.1136/jcp.53.7.502 on 1 July 2000. D

ownloaded from

nucleic acid. This development, backed byadvances in assay usefulness and reliabilityresulting from automation, might herald theend of the traditional stand alone clinical virol-ogy laboratory.

1 Baumforth KRN, Nelson PN, Digby JE, et al. The polymer-ase chain reaction. J Clin Pathol: Mol Pathol 1999;52:112–10.

2 Becker Y, Darai G. PCR: protocols for diagnosis of human andanimal virus diseases. Springer Lab Manual. Berlin:Springer-Verlag, 1995.

3 Hawkins A, Davidson F, Simmonds P. Comparison of viralloads among individuals infected with hepatitis C virus(HCV) genotypes 1, 2, and 3 by quantiplex HCV RNAassay versions 1 and 2, Roche Monitor assay, and anin-house limiting dilution method. J Clin Microbiol1997;35:187–92.

4 Gao SJ, Moore PS. Molecular approaches to the identifica-tion of unculturable infectious agents. Emerg Infect Dis1997;2:159–67.

5 Choo QL, Kuo G, Weiner AJ, et al. Isolation of a cDNAclone derived from a blood-borne non-A, non-B viralhepatitis genome. Science 1989;244:359–62.

6 Alter HJ. Descartes before the horse: I clone, therefore I am:the hepatitis C virus in current perspective. Ann Intern Med1991;115:644–9.

7 Simons JN, Leary TP, Dawson GJ, et al. Isolation of novelvirus-like sequences associated with human hepatitis. NatMed 1995;1:564–9.

8 Alter HJ, Nakatsuji Y, Melpolder J, et al. The incidence oftranfusion-associated hepatitis G virus infection and itsrelation to liver disease. N Engl J Med 1997;336:747–54.

9 Moore PS, Gao SJ, Dominguez G, et al. Primary characteri-sation of a herpesvirus agent associated with Kaposi’ssarcomae. J Virol 1996;70:549–58.

10 Strathdee SA, Veugelers PJ, Moore PS. The epidemiology ofHIV-associated Kaposi’s sarcoma: the unravelling mystery.AIDS 1996;10(suppl A):S51–7.

11 Renwick N, Halaby T, Weverling GJ, et al. Seroconversionfor human herpesvirus 8 during HIV infection is highlypredictive of Kaposi’s sarcoma. AIDS 1998;12:2481–8.

12 Di Alberti L, Piattelli A, Artese L, et al. Human herpesvirus8 variants in sarcoid tissues. Lancet 1997;350:1655–61.

13 Koerner K, Cardoso M, Dengler T, et al. Estimated risk oftransmission of hepatitis C virus by blood transfusion. VoxSang 1998;74:213–16.

14 Schachter J. DFA, EIA, PCR, LCR and other technologies:what tests should be used for diagnosis of chlamydia infec-tions? Immunol Invest 1997;26:157–61.

15 Rowley AH, Whitley RJ, Lakeman FD, et al. Rapid detectionof herpes simplex virus DNA in cerebrospinal fluid ofpatients with herpes simplex encephalitis. Lancet 1990;335:440–1.

16 Aslanzadeh J, Osmon DR, Wilhelm MP, et al. A prospectivestudy of the polymerase chain reaction for the detection ofherpes simplex virus in cerebrospinal fluid submitted to theclinical virology laboratory. Mol Cell Probes 1992;6:367–73.

17 GuVond T, Dewilde A, Lobert P-E, et al. Significance andclinical relevance of the detection of herpes simplex virusDNA by the polymerase chain reaction in cerebrospinalfluid from patients with presumed encephalitis. Clin InfectDis 1994;18:744–9.

18 Lakeman FD, Whitley RJ, and the National Institute ofAllergy and Infectious Diseases Collaborative AntiviralStudy Group. Diagnosis of herpes simplex encephalitis:application of polymerase chain reaction to cerebrospinalfluid from brain-biopsied patients and correlation with dis-ease. J Infect Dis 1995;171:857–63.

19 Cinque P, Brytting M, Vago L, et al. Epstein-Barr virusDNA in cerebrospinal fluid from patients with AIDS-related primary lymphoma of the central nervous system.Lancet 1993;342:398–401.

20 Tanel RE, Kao SY, Niemiec TM, et al. Prospective compari-son of culture vs genome detection for diagnosis of entero-viral meningitis in childhood. Arch Pediatr Adolesc Med1996;150:919–24.

21 Thoren A, Widell A. PCR for the diagnosis of enteroviralmeningitis. Scand J Infect Dis 1994;26:249–54.

22 Yerly S, Gervaix A, Simonet V. Rapid and sensitive detectionof enteroviruses in specimens from patients with asepticmeningitis. J Clin Microbiol 1996;34:199–201.

23 Mitchell PS, Espy MJ, Smith TF, et al. Laboratory diagnosisof central nervous system infections with herpes simplexvirus by PCR performed with cerebrospinal fluid speci-mens. J Clin Microbiol 1997;35:2873–7.

24 Read SJ, JeVery KJM, Bangham CRM. Aseptic meningitisand encephalitis: the role of PCR in the diagnostic labora-tory. J Clin Microbiol 1997;35:691–6.

25 Echevarria JM, Casas I, Tenorio T, et al. Detection ofvaricella-zoster virus-specific DNA sequences in cerebro-spinal fluid from patients with acute aseptic meningitis andno cutaneous lesions. J Med Virol 1994;43:331–5.

26 JeVery KJM, Read SJ, Peto TEA, et al. Diagnosis of viralinfections of the central nervous system: clinical interpret-ation of PCR results. Lancet 1997;349:313–17.

27 Schlesinger Y, Tebas P, Gaudreault-Keener M, et al. Herpessimplex virus type 2 meningitis in the absence of genitallesions: improved recognition with use of the polymerasechain reaction. Clin Infect Dis 1995;20:842–8.

28 Tedder DG, Ashley R, Tyler KL, et al. Herpes simplex virusinfection as a cause of benign recurrent lymphocytic men-ingitis. Ann Intern Med 1994;121:334–8.

29 Read SJ, Kurtz JB. Laboratory diagnosis of common viralinfections of the central nervous system by using a singlemultiplex PCR screening assay. J Clin Microbiol 1999;37:1352–5.

30 The TH, van der Bij W, van den Berg AP, et al. Cytomega-lovirus antigenaemia. Rev Infect Dis 1990;12(suppl 7):S734–44.

31 Wolf DG, Spector SA. Early diagnosis of human cytomega-lovirus disease in transplant recipients by DNA amplifica-tion in plasma. Transplantation 1993;56:330–4.

32 Burgart LJ, Heller MJ, Reznicek MJ, et al. Cytomegalovirusdetection in bone marrow transplant patients withidiopathic pneumonitis. A clinicopathological study of theclinical utility of the polymerase chain reaction. Am J ClinPathol 1991;96:572–6.

33 Evans PC, Soin A, Wreghitt TG, et al. Qualitative and semi-quantitative polymerase chain reaction testing for cytome-galovirus DNA in serum allows prediction of CMV relateddisease in liver transplant recipients. J Clin Pathol 1998;51:914–21.

34 Picciotto A, Campo N, Brizzolara R, et al. HCV-RNA levelsplay an important role independently of genotype inpredicting response to interferon therapy. Eur J Gastroen-terol Hepatol 1997;9:67–9.

35 Shiratori Y, Kato N, Yokosuka O, et al. Predictors of the eY-cacy of interferon therapy in chronic hepatitis C virusinfection. Gastroenterology 1997;113:558–66.

36 Ho DD, Neumann AU, Perelson AS, et al. Rapid turnover ofplasma virions and CD4 lymphocytes in HIV-1 infection.Nature 1995;373:123–6.

37 British HIV Association. Guidelines for antiretroviral treat-ment of HIV seropositive individuals. Lancet 1997;349:1086–92.

38 Mellors JW, Rinaldo CR, Jr, Gupta P, et al. Prognosis inHIV-1 infection predicted by the quantity of virus inplasma. Science 1996;272:1167–70.

39 Bell H, Hellum K, Harthug S, et al. Genotype, viral load andage as independent predictors of treatment outcome ofinterferon-alpha 2a treatment in patients with chronichepatitis C. Scand J Infect Dis 1997;29:17–22.

40 Mutimer D, Pillay D, Dragon E, et al. High pre-treatmentserum hepatitis B virus titre predicts failure of lamivudineprophylaxis and grafts re-infection after liver transplanta-tion. J Hepatol 1999;30:715–21.

41 Gerhold D, Rushmore T, Caskey CT. DNA chips:promising toys have become powerful tools. Trends BiochemSci 1999;24:168–73.

42 Lipshutz RJ, Fodor SP, Gingeras TR, et al. High densitysynthetic oligonucleotide arrays. Nat Genet 1999;21(suppl1):20–4.

43 Head SR, Parikh K, Rodgers YH, et al. Solid-phasesequence scanning for drug resistance detection intuberculosis. Mol Cell Probes 1999;13:81–7.

506 Read, Burnett, Fink

on 29 July 2018 by guest. Protected by copyright.

http://jcp.bmj.com

/J C

lin Pathol: first published as 10.1136/jcp.53.7.502 on 1 July 2000. D

ownloaded from