Testing for viruses – the way forward with PCR James Lowther

Viruses in shellfish; the problems Viruses (esp. norovirus and hepatitis A virus) are recognised as the principle cause of illness following bivalve shellfish consumption worldwide No known recent outbreaks of HAV following consumption of British shellfish Several outbreaks or incidents of norovirus food poisoning related to shellfish consumption reported annually Occasional large outbreaks may cause significant economic impact on producers No viral standards for shellfish in UK or EU legislation; requirement for robust fit-for purpose tests AND more data on background and outbreak events

Detection of norovirus in shellfish Straub et al. reported culture of human norovirus from faeces in 2007; other labs unable to reproduce results No realistic short- to mid-term prospect of culture methods for shellfish Detection of virus particles using microscopy or virus coat proteins using immunology (e.g. ELISA) currently lacks sensitivity required for low virus levels in shellfish – no reports of successful application to “real” food samples Identification of virus in shellfish currently only possible through “molecular methods” - specific detection of viral nucleic acid (PCR)

Real-time PCR (TaqMan) In last 5 years most labs worldwide have replaced “conventional” PCR-based methods for norovirus detection with real-time PCR (TaqMan) -based methods TaqMan technology enables quantitative analysis of data Simultaneous amplification and detection of virus sequence allows more rapid sample throughput In-built probe confirmation eliminates necessity for costly sequencing-based confirmation of positive PCR results

Principles of TaqMan Fluorescence released by TaqMan probe is proportional to the amount of PCR product amplified Fluorescence data is collected at every cycle by the real-time PCR machine The threshold cycle (Ct value), at which the fluorescence rises above a specified threshold level is identified for each reaction by the detection software The greater the initial amount of target DNA, the sooner exponential amplification is detected, and the lower the Ct value will be

Principles of TaqMan Most concentrated Least concentrated

Principles of TaqMan

Principles of TaqMan

Variety of virus detection methods 23 international labs involved in 2006 ring trial organised by Cefas as European Community reference lab (detection of norovirus in contaminated oysters) Virus extraction; 13 methods – most targeting digestive gland Viral RNA extraction – 29 methods RT-PCR - one and two-step, conventional single round, nested and semi-nested and real-time RT-PCR formats used Primers/probes – at least 13 different sets Development of standardised methodology desirable

European efforts towards development of a standard method for detection of viruses in food: CEN/TC275/WG6/TAG4 CEN is ISO equivalent body CEN/TC 275– Food analysis – Horizontal methods WG 6 – Microbial contamination TAG 4 – Detection of viruses in food Working group comprised of European food and water virology experts, circa 30 participants from 12 countries

CEN standard method TAG4 will circulate draft standard to parent bodies within next 6 months Hope to receive funding for full characterisation/validation of method soon Standard stipulates:- Treatment of 2g digestive glands with proteinase K solution followed by extraction of viral RNA with silica (recommends use of Nuclisens magnetic silica technology) Use of “one-step” TaqMan RT-PCR (recommends Invitrogen kit) Stringent criteria for primer selection (recommends particular sets) Use of particular classes of controls; positive and negative process controls, positive and negative RT-PCR controls etc. Method for quantification and expression of test results Many labs both within and outside EU (Australia, NZ, Peru) adopting or preparing to adopt CEN procedure

Prevalence Results of Cefas studies using TaqMan detection have detected norovirus RNA in 45-68% of samples Similar results in equivalent studies in e.g. France, Ireland, Japan Legal standards based on absence of detectable norovirus RNA is most simple approach but likely to be problematic (cf. Singapore, HK import requirements) Major industry impact *may* overestimate risk (further data required) Quantitative element of testing very important Necessity for proper systematic information on background levels and levels in outbreaks for accurate assessment of risk

Prevalence Unit of measurement stipulated in CEN standard will be detectable virus genome copies/g digestive gland (cf. Ct values, PCR units) Limit of detection method dependent but ~15 copies/g Since adoption of approach (Mar 08) levels detected at Cefas range from LOD to 1500 copies/g Based on previous data levels up to 3000 copies/g may be found routinely during winter Levels >5,000 copies/g are more unusual Note: these are preliminary estimates only – may change as we generate more data

Seasonality All studies at Cefas demonstrate distinct seasonality both in terms of prevalence and levels

Seasonality

Seasonality Possible causes of seasonality Increased levels of norovirus in community during winter; “winter-vomiting disease” Increased persistence of virus particles at lower temperatures and lower solar irradiation levels Decreased clearance of virus by shellfish due to lower metabolism during winter Observed seasonality consistent with predominance of shellfish-related outbreaks in winter period Caution!! Shellfish sampled in summer can return very high results e.g. ~10,000 copies/g from oysters harvested end of June

Correlation with illness Data on virus levels in shellfish causing illness incomplete Shellfish from relevant batch available/collected for testing infrequently following illness incidents/outbreaks Where “from the restaurant” shellfish have been available, sample sometimes norovirus positive at reasonably high levels, sometimes positive at low levels or negative Cause of illness (toxin, virus, other pathogen) not always established Epidemiological link with shellfish consumption not always confirmed Requirement for more systemic epidemiological data

Correlation with illness “Dubious” post-harvest practises can increase risk of illness (not in Scotland!!) Outbreak Malta, Jan 2007:- 50+ people ill with gastroenteritis (21 faecal samples norovirus positive) following consumption of oysters (grown in France) at 2 hotels, 1 private home Cefas supplied with 2 samples, 1 from hotel, 1 from importer Norovirus sequence match between oyster and faecal samples Shellfish probably stored in contaminated water by the importer prior to sale

Correlation with illness Shellfish contaminated with low levels of norovirus CAN pose risk of infection Outbreak Cardiff, Dec 2004:- 6 people ill with gastroenteritis following consumption of oysters (grown in Netherlands) at restaurant Oysters from restaurant tested, positive for norovirus at low levels ~LOD Sequence matches for both GI and GII norovirus (unusual strains) found between oysters and faecal sample Strong evidence for oysters as route of infection

Correlation with illness Cefas study with industry collaborators to systematically correlate reports of illness with norovirus levels in oysters Shellfish supplied to collaborating restaurants during winter 2006-2007 tested for norovirus at Cefas Customer reports of illness following oyster consumption screened for similarity of symptoms, time of onset to norovirus Using restaurant and supplier traceability, illness reports linked to particular batches of oysters Correlation with test results analysed

Correlation with illness Majority of samples (59%) and batches (74%) tested positive for norovirus Levels generally low, one batch provided exceptionally high results for norovirus GII Majority of batches (77%) NOT linked to illness reports No illness reports for batches where norovirus not detected

Correlation with illness “Attack rates” calculated for each batch associated with illness One batch (high norovirus results) linked to illness in 70+ people (21 incidents) – calculated attack rate ~4% (likely underestimate) Other incidents reported to local health authorities – several norovirus positive results for faecal samples Maximum attack rate for other batches ~1%, maximum number of linked incidents 4 (12 people)

Correlation with illness Positive correlation between attack rate and detected norovirus levels

Correlation with illness Correlation less clear cut at lower norovirus levels

Correlation with illness Conclusions of industry collaboration No detection of norovirus RNA = low health risk Detection of “low” levels = may be some risk, but unlikely to cause large outbreaks? Detection of “high” levels = risk likely to be severe Detection of “medium” levels = ??? For more effective risk management more data is required

Why not more illness reported?? Infectious dose of norovirus often quoted as <10 virus particles Most samples of shellfish tested contain >15 copies/g, often many times more Why are illness reports relatively scarce?? Discrepancy between norovirus RNA copies and viable virus particles? Strain-related effects?, e.g. GI norovirus as common as GII in SF, GII causes more than 90% of outbreaks Host-related effects. Blood group confers sensitivity to certain strains only? Under-reporting? – e.g. HPA IID study suggests 1:1500

Future work Improvements/refinements to methodology Validation of standard method Build up data on background levels through surveillance Continue to monitor harvesting areas in Scotland at 1-3 monthly intervals as part of FSAS-funded sanitary surveys Upcoming UK-wide long-term surveillance of 50 sites funded by FSA Continue to generate data on health risks of norovirus levels Data from sporadic illness outbreaks etc. Further industry collaborations?? Volunteer studies……

Acknowledgements The CEN/TC275/WG6/TAG4 The Cefas molecular virology team Industry collaborators