76 J Clin Pathol 2000;53:76–83

Quantitative molecular in patient J Clin Pathol: first published as 10.1136/jcp.53.1.76 on 1 January 2000. Downloaded from management

Wolfgang Preiser, Bianca Elzinger, Nicola S Brink

The detection of viral genome (that is, viral and speculate on the possible future role of nucleic acid, either DNA or RNA) has gained quantitative genomic testing for viral infec- enormous importance over the last decade in tions. the diagnosis and management of viral infec- tions and diseases (table 1). Nucleic acid testing (NAT) has proven its superiority over Detection and quantification of viral more conventional laboratory techniques in genome: general considerations several areas, for example: In principle, molecular biological assays that + for that cannot be cultured in test for nucleic acid consist of three compo- conventional cell culture systems or those nents: that are hazardous to isolate; + “front end”: sample preparation, often + for clinical sample material of small encompassing extraction and purification of volume or containing only small amounts of nucleic acids from the clinical material; ; + “middle part”: hybridisation to the target nucleic acid sequence; with genome amplifi- + where antibody seroconversion is delayed after an acute ; cation techniques this is where multiplica- tion of the target sequence occurs; + in the immunocompromised patient who “back end”: detection and possible quantifi- may have a suboptimal antibody response; + cation of target nucleic acid sequence; signal for the diagnosis of congenital or perinatally + enhancement techniques use this step for acquired viral . signal amplification. As well as qualitative analysis, the characteri- One of the most important objectives of the sation of viruses by analysis of the amplified front end is the removal of inhibitors of the region of the genome allows the detection and subsequent steps; this is particularly important identification of viral types or subtypes and with amplification techniques employing enzy- mutants. This is now used to monitor patients matic reactions which may be inhibited by on highly active antiretroviral therapy substances such as haemoglobin, heparin, and (HAART), based on the multitude of resist- bilirubin. Another important feature is the http://jcp.bmj.com/ ance associated mutations that have been char- eYciency of nucleic acid recovery. Some acterised in HIV infected patients receiving protocols “enrich” nucleic acid before 1 antiretroviral treatment. This allows a more extraction—for example, by centrifugation of rational therapeutic approach. Sequence analy- the specimen to concentrate its virus content. sis has also proved to be a valuable epidemio- The front end must be chosen according to the logical tool—for example, to investigate possi- type of clinical specimen (whole blood, ble links between cases of infection, as in plasma, urine, and so on), its quantity and 2 outbreaks of hepatitis B and hepatitis C quality, and must also take into account the on September 28, 2021 by guest. Protected copyright. infection.3 type of viral nucleic acid to be detected (DNA In addition to qualitative analysis and or RNA) and its location (intracellular or genome sequencing, methods have been devel- extracellular). In the context of introducing oped for the quantification of viral genomes. molecular techniques into a routine diagnostic While initially largely applied in the research service, ease of handling, speed, number of field, quantitative NAT has recently been samples to be extracted per run, reproduc- introduced into routine diagnostic virology. In ibility, and cost are obviously important this review, we will briefly introduce the diVer- considerations. The traditional method of ent methods available for viral genome quanti- phenol-chloroform extraction has in some fication, define their place in clinical virology, instances been replaced by faster and more Department of easily handled methods, for example those Virology, Royal Free based on nucleic acid binding to silica. and University College Table 1 Detection of viral genome (qualitative): clinical Following sample preparation and extraction applications Medical School and of the nucleic acid, diVerent techniques are University College + Detection of viruses that cannot easily be cultured: for available for detection and quantification of the London Hospitals NHS example, hepatitis B, human papillomavirus, HIV,hepatitis C target DNA or RNA. Quantification tech- Trust, Windeyer + Where maximum sensitivity for viral detection is required: niques include: Building, 46 Cleveland for example, ocular, amniotic, cerebrospinal fluid for presence of Street, London , varicella-zoster, JC virus, cytomegalovirus, and + target amplification techniques, for example W1P 6DB, UK rubella virus polymerase chain reaction based assays and W Preiser + Where antibody testing is inappropriate: for example, in a nucleic acid sequence based amplification primary infection (before antibody seroconversion; where B Elzinger (NASBA); and NSBrink passively acquired antibody may complicate the diagnosis in babies born to HCV or HIV infected mothers or in + signal amplification techniques, for example immunosuppressed patients branched DNA assays and hybrid capture Correspondence to: + To exclude infectivity: for example of blood and blood products Dr Brink assays. Quantitative molecular virology 77

AP J Clin Pathol: first published as 10.1136/jcp.53.1.76 on 1 January 2000. Downloaded from PCR products Na+ OH– AP probe

"capture" denaturing hybridisation

B B

SA SA AP

black 96 well microtitre plate

B

LIGHT SA photons per second

addition of Dioxetane substrate Figure 1 Principle of the enzyme linked assay (ELONA) for quantification of PCR products. AP,alkaline phosphatase; B, biotin; PCR, polymerase chain reaction; SA, streptavidin. Quantification of viral genome: specific the same test run with the samples to be methods analysed and may be “external” or “internal” TARGET AMPLIFICATION: POLYMERASE CHAIN standards, depending on whether they undergo REACTION amplification in separate reaction tubes or Methods that amplify the target sequence rely whether they are added to each sample in the on enzymatic actions to increase specifically same test tubes. A combination of internal and the number of target sequence copies in the external calibrators may also be used. Com- sample. This requires a careful laboratory set petitive PCR is based on the comparison of the up and stringent practices to prevent contami- amounts of unknown target sequence pro- nation by carryover of the end product duced in relation to a known quantity of a (containing possibly several million times the competitive template.4 Competitive PCR based http://jcp.bmj.com/ original amount of target sequence) into other quantification assays have been developed for a samples or reagents, causing false positive variety of viruses including cytomegalovirus, results. Kaposi’s sarcoma associated herpesvirus The most widely used technique is the (HHV-8),5 and hepatitis C virus (HCV).6 polymerase chain reaction (PCR). Briefly, a Methods of signal detection play an integral DNA target sequence—in the case of RNA role in the accuracy of PCR based quantifica- viruses (or viral messenger RNA) the RNA is tion assays. The optimal time for signal detec- first reverse transcribed into complementary tion is to measure the amount of amplicon on September 28, 2021 by guest. Protected copyright. DNA (reverse transcription or RT-PCR)—is generated before the “plateau” phase of ampli- amplified exponentially by a series of duplicat- fication is reached—that is, while the reaction is ing reactions. Methods of genome quantifica- still in its exponential phase and the amplicon tion using PCR based assays include: concentration proportional to the amount of + limiting end point dilutions of the PCR target sequence at the start. This determines products; the range over which accurate quantification + amplification with external standards; can be achieved (“dynamic range”, mostly

+ amplification with internal standards; expressed in log10 steps). Detection is often + amplification using a combination of inter- achieved in a microtitre plate format by means nal and external standards; of specific binding of enzyme labelled reagents + competitive amplification methods.4 to the amplified sequence; these can be Quantification of viral genome by “limit complementary nucleic acid sequences dilution” measures the amount of target (probes), anti-nucleic acid antibodies or a sequence present in a sample by subjecting it at biotin–streptavidin system. Finally, an enzy- diVerent dilutions to PCR and then detecting matic reaction leads to a quantifiable colori- the presence of amplicons by electrophoresis metric or chemiluminescent signal (analogous on an ethidium bromide stained agarose gel. to the enzyme linked immunosorbent assay This method of viral genome quantification is (ELISA) for the detection of antigen-antibody labour intensive and slow. Quantification may reactions).7 This is illustrated in fig 1. also be achieved by means of standards Recent additions to quantitative PCR meth- (calibrators) of known concentration, allowing ods are the LightCycler™ technology (Idaho the construction of a calibration curve against Technologies Inc, Roche Molecular Biochemi- which the signal obtained from the samples is cals) and TaqMan® chemistry (Perkin Elmer plotted. The calibrators are normally run on Biosystems), combining rapid thermal cycling 78 Preiser, Elzinger, Brink

for PCR and real time detection of the reaction Because the resulting double stranded DNA

kinetics by means of fluorimetry. carries the additional T7 promoter sequence its J Clin Pathol: first published as 10.1136/jcp.53.1.76 on 1 January 2000. Downloaded from With LightCycler, there are several possibili- sense strand can serve as a template for T7 ties for kinetic rather than end point DNA RNA polymerase for synthesis of an RNA quantification: a DNA binding dye (for exam- strand. The reaction then enters the amplifica- ple, SYBR Green I) binding to the newly syn- tion phase during which, in a cyclic manner, thesised double stranded DNA and thereby the newly formed RNA acts as a template for generating a fluorescent signal proportional to the to synthesise DNA. It its concentration, with DNA melting curve is then digested again by the RNAse H and analysis allowing confirmation of the identity of replaced by a complementary DNA strand the amplicon; or two hybridisation probes which allows the T7 RNA polymerase to binding to adjacent sequences within the produce more RNA molecules entering the amplicon leaving one base in between, thus amplification cycle anew. Three internal cali- allowing fluorescence resonance energy trans- brators that are coextracted and coamplified fer (FRET) between the donor fluor on the 3' (Qa, Qb, and Qc) allow for quantification. The end of one probe and the acceptor fluor on the quantity of amplified target genome and 5' end of the other one. calibrators is measured using electrochemilu- With TaqMan, a hybridisation probe binds minescence, and the signal produced by the to the amplicon during PCR but is degraded target genome is compared with the signals by the nuclease activity of the Taq polymerase from the three calibrators.4 NASBA is mar- extending the primer; this increases the keted by Organon Teknika as part of the distance between the quencher and the NucliSens system for the quantitative detection reporter dyes with which the probe is labelled, of HIV RNA (NucliSens HIV-1 QT). The so that the emission energy of the reporter assay has a dynamic range of 4 logs and could dye is no longer transferred to the quencher initially detect down to 400 HIV-1 RNA and the fluorescent emission intensity in- copies/ml. More recently a more sensitive assay creases. has been developed which can detect down to Quantitative PCR assays are commercially 40 HIV-1 RNA copies/ml. available for cell-free HIV (HIV RNA in plasma), hepatitis C (HCV RNA in serum or SIGNAL AMPLIFICATION: HYBRIDISATION plasma) and B viruses (HBV DNA in serum or TECHNIQUES plasma), as well as for cytomegalovirus (CMV Hybridisation procedures are based on the DNA in leucocytes or plasma) (Amplicor detection of a single stranded (that is, dena- HIV-1, HCV, HBV, CMV Monitor™, Roche tured) nucleic acid sequence by means of a Diagnostics). The use of the nucleotide dUTP labelled nucleic acid probe. This probe is com- together with the enzyme UNG (AmpErase®) plementary to the genome region in question helps to prevent contamination by amplicons. and therefore specifically binds to the target The HIV-1 and HCV monitor assays initially nucleic acid. This binding can then be http://jcp.bmj.com/ had a 4 log dynamic range able to detect down visualised, depending on the nature of the to 400 copies/ml of HIV-1 RNA and 2000 label, by autoradiography or an enzymatic copies/ml of HCV RNA. However, recent reaction, leading to a colour change. improvements in the assays have allowed for Quantitative hybrid capture™ systems (Di- the detection of as low as 20 copies/ml of gene Corporation; Murex) are commercially HIV-1 RNA and 600 copies/ml of HCV RNA . available for human cytomegalovirus DNA in Future developments include the automation leucocytes and hepatitis B virus (HBV) DNA of quantification assays. Roche Diagnostics in serum. These assays use RNA probes which on September 28, 2021 by guest. Protected copyright. have developed the Cobas Amplicor™ bench- bind to the target DNA from the clinical top analyser which incorporates thermal cy- specimen to form RNA:DNA hybrids. These cling, pipetting, incubation, washing, and are captured onto a solid phase coated with reading. anti-RNA:DNA hybrid capture antibodies. The same antibodies labelled with the enzyme TARGET AMPLIFICATION: NUCLEIC ACID alkaline phosphatase are then used to bind to SEQUENCE BASED AMPLIFICATION the solid phase bound RNA:DNA hybrids Nucleic acid sequence based amplification (“sandwich” style immunoassay). Finally, the (NASBA) is an isothermal RNA amplification amount of bound alkaline phosphatase is reaction which makes use of three diVerent measured quantitatively by addition of a enzymes active at 41°C: AMV-RT (avian chemiluminescent substrate and counting of myeloblastosis virus reverse transcriptase), photons emitted. Because several alkaline RNAse H, and T7 RNA polymerase. Primer 1, phosphatase molecules are conjugated to containing a T7 RNA polymerase promoter each anti-RNA:DNA hybrid antibody and sequence, anneals to the target sequence, single because multiple conjugated antibodies bind stranded RNA, and is used by the reverse tran- to each captured hybrid molecule, a 3000-fold scriptase to form an RNA:DNA hybrid. The or higher signal amplification is achieved RNA strand is then degraded by the action of (fig 2). RNAse H and a second DNA strand synthe- sised by extension of primer 2 by the reverse SIGNAL AMPLIFICATION TECHNIQUES: BRANCH transcriptase. If the target sequence is DNA, DNA ASSAY this is heat denatured to allow attachment of With branch DNA (bDNA) assays (Quan- primer 1 and synthesis of a double stranded tiplex®, Chiron Diagnostics) a combination of DNA carrying the T7 promoter sequence. synthetic oligonucleotide probes binds to and Quantitative molecular virology 79

Substrate J Clin Pathol: first published as 10.1136/jcp.53.1.76 on 1 January 2000. Downloaded from

Chemiluminescent Detector antibody, signal alkaline phosphatase conjugated

RNA:DNA hybrid Viral DNA

capture antibody

Solid phase Figure 2 Signal amplification: hybrid capture assay.

measures the amount of viral RNA or DNA in Table 2 Quantification of viral genome: clinical the sample. Target viral nucleic acid is captured applications

onto a solid phase by means of oligonucleotide Prognostic marker: for example, disease progression in HIV probes. Several other species of probes then Therapeutic marker: for example, for the initiation and monitoring bind to the target sequence and to each other in antiviral treatment To assess the risk of infection: for example, the risk of perinatal a sequential fashion, including branched DNA transmission in babies born to HIV and HCV infected women (bDNA) amplifier probes. Finally, multiple copies of the alkaline phosphatase labelled Clinical applications of viral nucleic acid probe hybridise to each amplifier probe on the quantification immobilised complex; addition of a chemi- The clinical applications of viral nucleic acid luminescent substrate cleaved by alkaline quantification are summarised in table 2. phosphatase then leads to light emission proportional to the amount of target sequence PROGNOSTIC MARKERS AND CLINICAL DISEASE in the sample (fig 3). ASSOCIATIONS ® Currently available are Quantiplex bDNA Viral genome quantification as a prognostic assays for the quantification of HIV-1 RNA, marker or an indicator of disease severity is HCV RNA, and HBV DNA (Chiron Corpora- now well established for certain viral infections, tion). Initially the HIV-1 bDNA assay had a in particular HIV. A study in the late 1980s sensitivity of 500 copies/ml, but more recently showed an association between the levels of an ultrasensitive version of the HIV assay has plasma HIV (as measured by end point http://jcp.bmj.com/ been produced which employs centrifugation dilution cultures) and disease stage, with indi- to concentrate virions in the plasma before viduals with AIDS having higher levels of tissue hybridisation and detects down to 50 copies/ culture infective doses of HIV per ml plasma.8 ml. Although culture based, this study provided the

Substrate Substrate on September 28, 2021 by guest. Protected copyright.

Chemiluminescent Chemiluminescent signal signal Label probe

Amplifier probe

Preamplifier probe HIV RNA Target probe

Capture probe (solution)

Capture probe (microwell)

Solid phase Figure 3 Signal amplification: branched DNA assay. 80 Preiser, Elzinger, Brink

first insight into the potential use of transplant recipients.15 A similar association as a prognostic marker. The association of was found in a group of bone marrow J Clin Pathol: first published as 10.1136/jcp.53.1.76 on 1 January 2000. Downloaded from plasma viral load with disease stage was transplant recipients, where only cytomegalovi- confirmed in the early 1990s in a study using rus viral load was a risk factor for the develop- the direct quantification of plasma RNA, using ment of cytomegalovirus associated disease.16 9 a PCR based method. More recently the dem- In HIV infected patients it has also been shown onstration of low levels of HIV-1 in the major- that median cytomegalovirus viral load is ity of long term non-progressors (defined as significantly greater in patients who subse- HIV infected individuals who remain healthy quently develop clinical disease.17 and immunologically normal for more than 10 It has been suggested that the amount of years) has provided further information on the cytomegalovirus in the urine from an infected role of HIV viral load as a prognostic marker.10 neonate may be directly related to the progno- It has also been shown that the post- 4 seroconversion viral load may predict progres- sis of congenital cytomegalovirus infection. It sion to AIDS.11 This study showed that a high is possible that this may be extended to the plasma HIV-1 RNA level (100 000 copies/ml) quantification of cytomegalovirus DNA in was the most powerful predictor of AIDS. In amniotic fluid as a predictor of severe fetal dis- contrast, a plasma HIV-1 RNA level of less ease in the future. than 10 000 copies/ml was not associated with Determination of viral load may also be use- progression to AIDS. Table 3 shows the clear ful for the diagnosis of disease associated with relation between viral load (serum HIV-1 other herpesviruses. Post-transplant lympho- RNA) and disease outcome in HIV infected proliferative disorder (PTLD) is a potentially individuals.12 It has also been found that life threatening complication following allo- plasma HIV-1 RNA levels are predictive of the geneic transplantation associated with infec- subsequent clinical course in perinatally in- tion with the Epstein-Barr virus (EBV). Recent fected infants.13 studies have linked EBV viral load in plasma Cytomegalovirus infection is common in and peripheral blood lymphocytes to the risk of immunosuppressed patients and may result in developing PTLD.18 19 This may be important severe disease. Rapid laboratory diagnosis of in monitoring patients at risk of disease devel- infection and the ability to predict which opment, particularly as the current method of patients are at risk of progression to clinical diagnosing PTLD is based on a combination of disease is therefore important. Qualitative clinical and histopathological features, an cytomegalovirus DNA detection has been used approach which may be too slow to be clinically in post-transplant monitoring programmes, useful.4 Determination of EBV viral load may but will inevitably involve the unnecessary treatment of certain patients if pre-emptive therefore be a useful marker to predict disease anti-CMV therapy is used (that is, treatment of development, thereby influencing patient mor- patients with evidence of cytomegalovirus bidity and possibly mortality, as management http://jcp.bmj.com/ infection but not disease). A study analysing of PTLD includes the reduction of immuno- 248 blood samples from immunocompromised suppressive drugs. Assays for the quantification patients showed that the positive predictive of human herpesvirus 6 and 7 have also been value of qualitative PCR positivity in the blood developed and it is possible that these will help with respect to the development of cytomega- define the pathogenic role of these ubiquitous loviral disease was 0.62.14 The initiation of pre- viruses.20 21 emptive anti-CMV therapy on the basis of the In contrast to the demonstrable value of qualitative detection of cytomegalovirus DNA HIV-1 and cytomegalovirus viral load as a on September 28, 2021 by guest. Protected copyright. in blood therefore results in the exposure of prognostic marker in HIV infected individuals, patients to potentially toxic drugs without no reliable association has been shown between clinical need. Refinement of cytomegalovirus serum HCV titres and the severity of liver surveillance after transplantation may be disease.22 For hepatitis B, however, a rough achieved by the monitoring the cytomegalovi- correlation between the presence of liver rus load, and this is likely to form an integral disease and the amount of serum HBV DNA part of post-transplantation monitoring pro- has been demonstrated. A study showed that grammes in the future. Critical here is the patients who were hepatitis B surface antigen determination of an appropriate cut o for V (HBsAg) and hepatitis B e antigen (HBeAg) therapeutic intervention. Using a PCR based positive all had detectable HBV DNA by PCR. quantification assay, Fox and colleagues A majority of these patients also had HBV showed that high levels of cytomegalovirus correlated with clinically apparent cytomegalo- DNA detectable by the less sensitive dot blot virus associated disease in a cohort of renal hybridisation and most had significant liver disease. In contrast, patients who had detect- Table 3 Relation of plasma HIV viral load to survival able serum HBsAg but no HBeAg, and lower times in HIV infected individuals mean serum transaminase levels, usually did not have detectable HBV virus DNA by dot HIV viral load 23 (RNA copies/ml plasma) Median survival time blot hybridisation. It is therefore likely that the level of HBV DNA is a more accurate > 36 270 5.1 years 13 021–36 270 7.4 years marker of virus replication than HBeAg alone 4531–13 020 9.5 years and may be important in the study of the < 4530 > 10 years pathogenesis of HBeAg negative pre-core 24 (from Mellors et al, 199612) mutations. Quantitative molecular virology 81

INITIATION AND MONITORING OF ANTIVIRAL copies/ml of serum HCV RNA, whereas 12

TREATMENT months of treatment are recommended for J Clin Pathol: first published as 10.1136/jcp.53.1.76 on 1 January 2000. Downloaded from One of the most important clinical applications patients with viraemia in excess of two million for viral nucleic acid quantification is for the copies/ml. Monitoring the therapeutic re- initiation, monitoring, and modification of sponse is currently done by qualitative PCR, antiviral treatment. For HIV-1 the goal of although genome quantification is likely to play antiretroviral therapy is to reduce plasma an important role in the future.29 It has also HIV-1 RNA levels to the lowest level possible been suggested that the rate of decline in virae- for as long as possible. This is logical given the mia during the first few weeks on interferon interaction between HIV-1 replication and treatment can predict the therapeutic outcome CD4 lymphocyte destruction, which in turn in chronically infected patients.30 influences clinical outcome.25 The first ques- tion is, at what plasma HIV-1 RNA level should MARKERS OF INFECTIVITY treatment be initiated? It has been suggested The use of viral genome quantification as a that all patients with plasma HIV-1 RNA levels marker of infectivity is likely to play an impor- of > 30 000–50 000 copies/ml should be of- tant role in patient management in the future. fered antiretroviral treatment regardless of the Possible uses include the assessment of the risk state of their clinical disease, whereas for indi- of transmission of HCV or HIV from mother to viduals with plasma HIV-1 levels of > 5000 but child in the peripartum period, together with < 30 000 copies/ml the decision to treat should the risk of infection after occupational expo- take into account the CD4 count and clinical sure to a bloodborne virus. Plasma or serum disease stage.25 Following initiation of anti- levels of HIV and HCV are important determi- retroviral therapy, viral load should be used to nants of risk in the perinatal transmission from monitor the therapeutic response. At baseline mother to child. For hepatitis C, mother to this should consist of two measurements two to child transmission is more likely if the maternal four weeks apart and thereafter every three to serum HCV RNA concentration is greater than four months in a clinically stable patient. A 0.5 106 to 107 genomes/ml.31 A study comparing log or greater reduction of plasma RNA levels transmission rates from various centres showed is considered to be the minimum response that only two of 30 women who transmitted indicating an antiretroviral eVect. If HIV-1 infection to their infants had a viral load of less RNA levels return to (or to within 0.3–0.5 logs than 106 copies/ml.32 Coinfection with hepatitis of) pretreatment levels, this could signify the C and HIV also increases transmission rates of selection of a resistant virus.25 However, other HCV (from 3.7% in patients infected with causes of a rising viral load, such as poor drug hepatitis C alone to 15.5% in women coin- compliance or an acute opportunistic infec- fected with HIV in one study), probably tion, should be excluded.26 because of the increase in HCV RNA in the Molecular quantification of HBV virus DNA serum of coinfected women.33 Although a study has been used to predict a response to from Italy has suggested a lower rate of http://jcp.bmj.com/ interferon and to monitor interferon treatment transmission in babies born by caesaean in patients infected with HBV. High pretreat- section compared with vaginal delivery (6% v ment levels of virus DNA are predictive of a 32%),34 the optimum mode of delivery for the poor therapeutic response to interferon.4 Sev- woman infected with HCV has not yet been eral factors influence the therapeutic response defined. It is possible, therefore, that HCV viral to antiviral treatment in patients infected with load may help in the choice of delivery in the HCV. These include the virus genotype future. Although the risk of transmission of (patients infected with genotype 1 respond less HIV-1 from mother to child is also increased on September 28, 2021 by guest. Protected copyright. well than patients infected with genotypes 2 or with a higher viral load, a recent study could 3) and the viral load. A study by Garson et al not demonstrate a threshold value of virus load showed that individuals who had a sustained that discriminated between transmitters and response to antiviral treatment (defined as a non-transmitters. It also showed transmission complete normalisation of aminotransferase and non-transmission occurring over the entire levels within three months of starting inter- range of each of the quantification assays used, feron, and maintained throughout the 12 indicating the need for caution when consider- month follow up period) had mean pretreat- ing viral load as a risk factor for transmission of ment HCV RNA levels that were approxi- HIV-1 infected mothers.35 It is likely, therefore, mately 10-fold lower than in individuals who that the prevention of perinatally acquired HIV did not have a sustained response.27 This was infection will largely be based on the use of confirmed in a later study.28 In the consensus antiretroviral treatment, elective caesarean sec- statement produced following the 1999 EASL tion, and the avoidance of breast feeding. international consensus conference on hepati- The role of viral load in assessing the risk of tis C, it was noted that while patients with HIV transmission, both sexually and through higher levels of viraemia (more than two occupational exposure, needs to be investigated million copies/ml) are less likely to respond to further. A recent study examined the relation of therapy, the level of viraemia should not be viral load to heterosexual transmission of HIV used as a reason to withhold treatment. The in haemophiliac men and showed an associ- level of viraemia does, however, influence the ation of high viral load with transmission of duration of treatment in patients infected with infection. This may help in the counselling of HCV genotype 1. Here the current data haemophiliac couples in the future.36 Although suggest that six months of treatment are there is no prospective study determining the suYcient for patients with less than two million risk of infection with a bloodborne virus after 82 Preiser, Elzinger, Brink

occupational exposure (through needlestick 4 Hodinka RL. The clinical utility of viral quantification using molecular methods. 1998;10:25–47. Clin Diagn Virol J Clin Pathol: first published as 10.1136/jcp.53.1.76 on 1 January 2000. Downloaded from accidents or mucous membrane or non-intact 5 Lock MJ, GriYths PD, Emery VC. Development of a quan- skin exposure) it is reasonable to assume the titative polymerase chain reaction for human herpesvirus 8. J Virol Methods 1997;64:19–26. risk of infection is related to the quantity of 6 Ravaggi A, Zonaro A, Mazza C, et al. Quantitation of hepa- virus in the source material. This in turn is titis C RNA by competitive amplification of RNA from denatured serum and hybridisation on microtitre plates. J influenced by the disease stage of the patient; Clin Microbiol 1995;33:265–9. for example it has been shown that HIV-1 viral 7 Whitby K, Garson JA. Optimisation and evaluation of a quantitative chemiluminescent polymerase chain reaction load increases with progression to AIDS (see assay for hepatitis C RNA. J Virol Methods 1995;51:75–88. above). 8 Ho DD, Tarsem M, Masud A. Quantitation of human immunodeficiency virus type 1 in the blood of infected A possible future development is the use of persons. N Engl J Med 1989;321:1621–5. serum HBV levels to clear infected health care 9 Semple MG, Kaye S, Loveday C, et al. Direct measurement of viraemia in patients infected with HIV-1 and its relation- workers who are found to be “low infectivity” ship to disease progression and zidovudine therapy. JMed hepatitis B carriers (that is, their serum does Virol 1991:35:38–45 10 Cao Y, Qin L, Zhang L, et al. Virologic and immunologic not contain hepatitis B e antigen) for the per- characterisation of long-term survivors of human immuno- formance of exposure-prone procedures. Cur- deficiency virus type 1 infection. N Engl J Med 1995;332: 201–8. rent Department of Health guidelines in the 11 Mellors JW, Kingsley LA, Rinaldo CR, et al. Quantitation of United Kingdom state that health care work- HIV-1 RNA in plasma predicts outcome after seroconver- sion. Ann Intern Med 1995;122:573–9. ers who are HBV infected but are hepatitis B 12 Mellors JW, Rinaldo CR, Gupta P, et al. Prognosis in HIV-1 carriers of low infectivity can perform infection predicted by the quantity of virus in plasma. Sci- ence 1996;272:1167–70. exposure-prone procedures as long as they 13 Shearer WT, Quinn TC, LaRussa P, et al. Viral load and have not been involved in a transmission event disease progression in infants infected with human 37 immunodeficiency virus type 1. N Engl J Med 1997;336: from health care worker to patient. Despite 1337–42. this, transmission events still occur, and seven 14 Kidd IM, Fox JC, Pillay D, et al. Provision of prognostic information in immunocompromised patients by routine such events have occurred in the United King- application of polymerase chain reaction for cytomegalovi- dom in so called “low infectivity” hepatitis B rus. Transplantation 1993;56:867–71. 15 Fox JC, Kidd IM, GriYths PD, et al. Longitudinal analysis carriers (Tedder R, personal communication). of cytomegalovirus load in renal transplant recipients using Addition of HBV viral load determination is a quantitative polymerase chain reaction: correlation with likely to add an extra margin of safety, while disease. J Gen Virol 1995;76:309–19. 16 Gor D, Sabin C, Prentice HG, et al. Longitudinal allowing health care workers at a low risk of fluctuations in cytomegalovirus load in bone marrow trans- transmission to continue their surgical plant patients: relationship between peak virus load, donor/ recipient serostatus, acute GVHD and CMV disease. Bone career. Marrow Transplant 1998;21:597–605. 17 Bowen EF, Sabin CA, Wilson P, et al. Cytomegalovirus (CMV) viraemia detected by polymerase chain reaction identifies a group of HIV-positive patients at high risk of Conclusions CMV disease. AIDS 1997;11:889–93. 18 Riddler SA, Breinig MC, McKnight JLC. Increased levels of Viral genome quantification now forms an circulating Epstein-Barr virus (EBV)-infected lymphocytes integral part of the management of HIV and decreased EBV nuclear antigen antibody responses are associated with the development of posttransplant lympho- infected patients, providing pretreatment proliferative disease in solid organ transplant recipients. prognostic information as well as monitoring a Blood 1994;84:972–84. 19 Yamamoto M, Kimura H, Hironaka T, et al. Detection and http://jcp.bmj.com/ therapeutic response. The next few years are quantitation of virus DNA in plasma of patients with likely to see the expansion of the clinical appli- Epstein-Barr virus-associated diseases. J Clin Microbiol 1995;33:1765–8. cations of nucleic acid quantification, with its 20 Clark DA, Ait-Khalid M, Wheeler AC. Quantification of use in the monitoring of HBV and HCV human herpesvirus 6 in immunocompetent and post- infected patients on antiviral treatment. Blood mortem tissues from AIDS patients by OCR. J Gen Virol 1996;77:2271–5. cytomegalovirus DNA levels will also guide 21 Kidd IM, Clark DA, Ait-Khaled M, et al. Measurement of the use of pre-emptive anti-CMV treatment in human herpesvirus 7 load in peripheral blood and saliva of healthy subjects by quantitative polymerase chain reaction. patients following bone marrow transplanta- J Infect Dis 1996;174:396–401. on September 28, 2021 by guest. Protected copyright. tion, allowing the more rational use of 22 Ahmed MM, Mutimer DJ, Martin B, et al. Hepatitis C viral load, genotype and histological severity in patients with potentially myelosuppressive antiviral treat- bleeding disorders. Haemophilia 1999;5:49–55. ment. International standardisation of mo- 23 Baker BL, Di-Bisceglie AM, Kaneko S, et al. Determination of hepatitis B virus DNA in serum using the polymerase lecular quantification techniques will allow chain reaction: clinical significance and correlation with comparison of results, enabling consensus clinical and biochemical markers. Hepatology 1991;13:632– 6. policies for infectious disease management to 24 Berger A, Braner J, Doerr HW, et al. Quantification of viral be produced. Genome quantification may load: clinical relevance for human immunodeficiency virus, hepatitis B virus and hepatitis C virus infection. Intervirol- also provide an accurate assessment when ogy 1997;41:24–34. evaluating the risk of infection, for example in 25 Saag MS, Holodniy M, Kuritzkes DR, et al. HIV viral load markers in clinical practice. Nature Med 1996;2:625–9. the transmission of HCV from mother to 26 Churchill D, Weber J. Measurement of plasma HIV viral child and in the HBV infected health care load. J Infect 1999;38:147–50. 27 Garson JA, Brillanti S, Whitby K, et al. Analysis of clinical worker. and virological factors associated with response to alpha interferon therapy in chronic hepatitis C. J Virol Methods 1995;45:348–53. We thank Dr J Garson for his helpful comments and advice. 28 Garson JA, Uhnoo I, Whitby K, et al. Virological, biochemi- cal and histological eVects of human lymphoblastoid inter- feron in Swedish patients with chronic hepatitis C. J Viral 1 Hammond J, Calef C, Larder B, et al. Mutations in retroviral Hepatitis 1997;4:325–31. genes associated with drug resistance. In: Korber B, Kuiken 29 EASL International Consensus Conference on Hepatitis C. C, Foley B, et al,eds.Human and AIDS 1998: a Paris 26–28 February 1999. Consensus Statement. J Hepa- compilation and analysis of nucleic acid and amino acid tol 1999;30:956–61. sequences. Los Alamos: Theoretical Biology and Biophysics 30 Walsh KM, Good T, Cameron S, et al. Viral kinetics can Group T-10, 1998:III-36–79. predict early response to alpha-interferon in chronic hepa- 2 Hawkins AE, Zuckerman MA, Briggs M, et al. Hepatitis B titis C. Liver 1998;18:191–5. nucleotide sequence analysis: linking an outbreak of acute 31 Dusheiko GM, Khakoo S, Soni P, et al. A rational approach hepatitis B to contamination of a cryopreservation tank. J to the management of hepatitis C infection. BMJ Virol Methods 1996;60:81–8. 1996;312:357–64. 3 Smith DB, Lawlor E, Power J, et al. A second outbreak of 32 Thomas SL, Newell ML, Peckham CS, et al. A review of hepatitis C virus infection from anti-D immunoglobulin in hepatitis C (HCV) vertical transmission: risks of transmis- Ireland. Vox Sanguinis 1998;76:175–80. sion to infants born with and without HCV viraemia and Quantitative molecular virology 83

human immunodeficiency virus infection. Int J Epidemiol 35 Cao Y, Krogstad P, Korber BT, et al. Maternal HIV viral 1998;27:1208–17. load and vertical transmission of infection: the Ariel Project J Clin Pathol: first published as 10.1136/jcp.53.1.76 on 1 January 2000. Downloaded from 33 Tovo PA, Palomba E, Ferraris G, et al. Increased risk of for the prevention of HIV transmission from mother to maternal–infant hepatitis C transmission for women infant. Nature Med 1997;3:549–52. co-infected with human immunodeficiency virus type 1. 36 Ragni MV, Faruki H, Kingsley LA. Heterosexual HIV-1 Italian Study Group for HCV infection in Children. Clin transmission and viral load in hemophilic patients. J Acquir Infect Dis 1997;25:1121–4. Immune Defic Syndr Hum Retrovirol 1998;17:42–5. 34 Paccagnini S, Principi N, Massironi E, et al. Perinatal trans- 37 Health Service Guidelines. Protecting health care workers mission and manifestation of hepatitis C virus infection in and patients from hepatitis B. London: department of a high risk population. Pediatr Infect Dis J 1995;14:195–9. Health, 1993 [HSG(93)40:19931–3]. http://jcp.bmj.com/ on September 28, 2021 by guest. Protected copyright.