Navigating 21st Century Virology

Back in the 1950s and 60s, routine virology diagnostics in teaching hospitals involved virus isolation in cell lines and serology, using techniques such as complement fixation tests. By the 1970s and 80s, electron microscopy and immunofluorescence techniques were firmly planted at the forefront of testing. Then the huge discovery of HIV in the 1980s paved the way for the development of enzyme immunoassays. This was followed in the early 1990s by the introduction of the popular polymerase chain reaction (PCR) assays. Nowadays, the availability of several new molecular techniques, most of them automated, is a legacy of three decades of nucleic acid research. But given the vast range of options currently available, it’s unsurprisingly difficult for today’s clinical virologist to select the techniques and technologies to best suit their needs.

Bring on the next generation

Let us begin with a technique that broke new ground and opened up a whole new world of scientific discovery and knowledge: capillary electrophoresis-based Sanger sequencing. This practice-changing method has enabled scientists to elucidate genetic information from any given biological system for almost 40 years now, and it’s still routinely used by virology laboratories for HIV and hepatitis B virus (HBV) sequencing. However, Sanger sequencing has fairly substantial limitations in the human genomics field in terms of throughput, scalability, speed and resolution. In a bid to address these limitations, next-generation sequencing (NGS) technology came to the fore. NGS is now a catch-all term to describe a number of sequencing technologies, many of which hold potential in routine diagnosis (see Figure 1).

NGS applications within virology have been summarized as (1):

i) detection of unknown viral pathogens and discovery of novel viruses,
ii) detection of tumor viruses,
iii) characterization of the human virome,
iv) sequencing of full-length viral genome,
v) characterization of viral genome variability and viral quasispecies,
vi) monitoring antiviral drug resistance, epidemiology of viral infections and viral evolution and quality control of live-attenuated viral vaccines.

But with a diverse range of technologies, the question remains: how useful are they in routine diagnostic virology? Published results for each system show them to be largely interchangeable (1).

A collaborative effort across Public Health England (PHE) sites is in the process of validating NGS for HIV-1 genotypic resistance testing with a view to fully replacing Sanger sequencing with MiSeq (Illumina); however, this has not yet been published. As with other NGS studies, it is expected that the assay threshold could be drilled down to one percent of minority species compared with 10–15 percent using Sanger sequencing. But the clinical benefit of detecting minority variants by NGS need to be evaluated both in the short and long term (2).

Fortunately, the impact of NGS in diagnostic virology involving HIV-1 tropism, genotypic resistance in influenza, HBV, hepatitis C virus (HCV) and cytomegalovirus (CMV) has already been summarized (3). Though the authors noted benefits including “increased sensitivity and eventually cheaper antiviral resistance tests,” they cautioned, “there is a risk that low percentage minority variants may be over interpreted. This could result in antiviral drugs, which may have been effective, being possibly denied to patients if proper clinical validation studies are not performed.” Furthermore, numerous papers from around the world have reported on the use of NGS technology in virus outbreak monitoring and mapping quasispecies variations; for example, influenza virus, hepatitis A virus, norovirus, enterovirus, and ebola virus during the 2014 West African outbreak (4, 5, 6, 7, 8).

Despite numerous publications demonstrating the advantages of using NGS in virology, its uptake has been severely slowed by the lack of widespread collaborations between human/cancer geneticists and virologists; the cost of equipment; and the continued availability of the well-established Sanger sequencing method for routine nucleic acid sequencing. As a result, NGS has not really taken off in the diagnostic clinical virology field, yet (2, 3). But that could all change within months – once the cost- and clinical-effectiveness of the relevant technologies are established.

Automation of liquid handling

In the new millennium, reagent volumes have decreased, but testing workload has increased. It is now well recognized that manual pipetting is a liability that increases the likelihood of original sample contamination and mix-ups. Whether it is enzyme immunoassay or nucleic acid extraction, repetitive steps demand precision. So it’s understandable that the preferred option these days is to decrease manual repetitive tasks and automate, wherever possible.

Liquid handling for nucleic acid extraction has come a long way since the 1990s when manual mixing of PCR reagents was the norm. The diagnostic market is now flooded with liquid-handling machines, each generation being an improvement over the last, both in capacity and speed. They are also constantly evolving to accept different types of samples, process lower volumes suitable for generic and/or manufacturer’s own specific reagents, and dispense liquids into 96-well plate format/circular disc/other topographical variations. Most machines have a bi-directional interface with a laboratory information management system (LIMS) and some have evolved to become a complete solution macro-robot or miniaturized technology using microfluidics (see below). Roche, Qiagen, Tecan, Abbott, Siemens, Beckton Dickenson and Life Technologies are just some of the major manufacturers in this field.

Automated liquid handling technologies are clearly beneficial, but there are instances when manual methods are still preferred, specifically when dealing with precious low volume samples, like pediatric blood, vitreous fluid and CSF samples, which still need initial manual processing to prevent ‘dead volume’ loss in automation. As a general rule, when sample numbers are low in a laboratory, capital investment in automation equipment is not cost-effective; but as sample numbers increase, so does the return on investment.

 

The Rare and Imported Pathogens Laboratory (RIPL) in Porton Down, Microbiology Services Public Health England (PHE), UK, for example, has replaced all manual pipetting processes in indirect immunofluorescence assays with the IF Sprinter (Euroimmun), thus fully automating the process – from dilution and dispensing of samples to incubation and washing of microscope slides. The upgrade has resulted in increased capacity and decreased manual pipetting errors and inter-operator variations.

Further, the entire molecular virology setup, from nucleic acid extraction to detection, at South London Specialist Virology Centre, King’s College Hospital in London, has been automated since 2008. A constant challenge in virology is that one can never predict a sudden increase in workload, such as the 2009 pandemic influenza virus. During the initial few weeks of this pandemic, we were inundated with 800 respiratory samples per day for influenza A virus RNA testing, which stretched our automation to its limit.

In 2013, we installed the Freedom EVO platform (TECAN) to separate serum and plasma for the extraction of RNA and DNA and for liquid dispensing into various sample racks, plates and capillaries. Earlier variations of this platform are still in use and they perform the following tasks: sample aliquoting, sample dilution, complement fixation test (CFT) plates, agglutination assays, ELISAs and final aliquoting for sample storage. All samples are tracked via barcode, and all worksheets are linked to the LIMS. This has substantially improved capacity within the lab, as well as efficiency and the reproducibility of assays.

The latest in PCR and its rivals

Nucleic acid amplification tests have been the focus of much development since the mid-1990s because of the sensitivity, specificity, turnaround time (TAT) and contamination issues surrounding the now redundant block-based assays. Today in the commercial world, thermal processes like PCR compete with isothermal processes like transcription mediated amplification (TMA), loop mediated isothermal amplification (LAMP), strand displacement amplification (SDA), and so on, in terms of lab spend. Nevertheless, real-time PCR forms the backbone of molecular diagnostic virology processes worldwide. A simple PubMed search using the term “real time PCR virus” led to a listing of nearly 15,000 articles with the first article dated 1993 (9), and this is in the narrow field of virology (Figure 2). Since then, new developments in real-time PCR have yielded improved sensitivity, probe-based specificity, increased capacity to multiplex and differentiate (because of the availability of numerous fluorescent dyes – from six to 30 over the last decade), and increasing availability of analytical software. Today, assays can be developed rapidly and reagents can be purchased for as little as £1 per reaction.

Though impressive and extremely useful in virology testing, nucleic acid amplification sequences one or a small set of organisms to rapidly identify selected pathogens at the species or strain level, but cannot be multiplexed to the degree required to detect hundreds to thousands of different organisms. A downfall that has been addressed by the new Lawrence Livermore Microbial Detection Array (LLMDA) (10). LLMDA is a multiple displacement amplification (MDA)-based whole genome amplification, which uses Phi29 polymerase, known to offer high processivity and low error rate when compared with Taq polymerase. It can be used for whole transcriptome amplification with 2.1 million probes available representing different pathogens. The authors of the original research have now successfully used LLMDA to detect a range of emerging viruses, including dengue virus, West Nile virus, Japanese encephalitis virus, tick-borne encephalitis virus, yellow fever virus, to name but a few.

A wide variety of automated liquid handling platforms are available for running real-time PCR and other nucleic acid detection techniques, many of which can accommodate both commercial and in-house tests. Virologists have accepted and adapted real-time PCR, resulting in decades of accumulated collective experience and confidence in its results. It was only a matter of time before miniaturization and microfluidics took over, paving the way for the first ever fully integrated and automated nucleic acid sample preparation, amplification, and real-time detection system (11). Released by diagnostics firm Cepheid in 2007, the system consists of an instrument, a personal computer, and disposable fluidic cartridges. The ease of use of this instrument has led to its application as a molecular point-of-care test (POCT).

POCT on the up

There is now widespread use of POCTs in clinics and in the field – for example, antigen detection (influenza and respiratory syncytial viruses), antibody detection (HIV and HCV) and cell counts (hemoglobin, neutrophils and CD4 count). A great deal of research and development has focused on POCT in recent years and this trend is showing no signs of relenting. If anything, it is accelerating. Today, a number of commercial companies now operate in the molecular POCT sector, developing technologies that range from cartridge-, membrane-, and microarray-based solutions to isothermal and thermal molecular systems. Some are just entering into human clinical trials and others have tremendous potential. In 2001, the World Health Organization outlined the key – or so-called “ASSURED”  – criteria that each POCT must meet to be deemed viable (see Table 1) (12).

 

Some of the available technologies have been summarized in Table 2, with a new proposed rating based on this very practical ASSURED criteria. In this proposed rating, presence of each attribute within the ASSURED criteria gets a score of 1, with a maximum possible score of 7. For the sake of simplicity, in this article, it is assumed that all these molecular technologies are sensitive and specific enough to warrant a starting baseline score of 2 out of a maximum possible 7 in the ASSURED criteria.

The companies that now operate in this arena vary widely in their expertise and technological approach. Theranos, for example, has been a pretty secretive company, having not released a single photograph of its equipment or published any data in peer-reviewed journals (13, 14). The two papers published (as highlighted in Table 2) did not originate from the manufacturing company. However, the US-based tech firm has recently received a welcome boost by US FDA regulators who have been wooed by Theranos’ technology; the firm has been cleared to market its herpes simplex 1 virus IgG (HSV-1) nationally, and it’s also received a highly sought after Clinical Laboratory Improvement Amendments (CLIA) Waiver, permitting use of it in locations outside of traditional clinical laboratories. How did they manage it? According to their press statement, “Theranos provided study data from 818 subjects of varying age and ethnicity, demonstrating that its system could be run accurately using only a finger stick as well as a traditional venous draw across large numbers of Theranos devices, all compared against an FDA cleared, commercially available reference method”. Other manufacturers, like Tetracore (15), have ventured only recently into human pathogen detection, and like Theranos, there are some firms that have also produced an array of user-friendly products for the laboratory, which can also be used in the POCT setting (BioFire FilmArray and Genmark Dx eSensor, for example) (16, 17).

With regards to technological capabilities of these tests, some incorporate new PCR methods, like Extreme PCR, developed by The Wittwer DNA Lab in Utah University, which has a reaction time between 15 and 60 seconds (18). Others have adapted existing reagent technologies like Xtreme Chain Reaction and Biomeme, but using smartphone capabilities (19, 20). The number of PubMed citations listed in Table 2 is an indication of the maturity of the technology and its acceptance within the medical field. Though the list of POCTs in the table is by no means exhaustive, it is indicative of the technological developments taking place in this field; any one of these technology types has the potential to be a diagnostic game-changer.

It’s all going digital

As with many fields of pathology, digitization is now revolutionizing the field of PCR (21). This latest refinement of conventional PCR can be used to directly quantify and clonally amplify nucleic acids. Like PCR, digital PCR carries out one reaction per single sample; however, the sample is separated into a large number of partitions – either as a series of droplets or split into nanoscale reaction wells – and the reaction is carried out in each partition individually. The outcome? High sensitivity and precision, high tolerance to inhibitors and amenability to quantitation. This allows a more reliable quantitation of nucleic acids like CMV quantitation standards (22). Research has been published on its use in Chlamydia trachomatis detection in trachoma cases, as well as identification of rotavirus in water, HIV-1 proviral DNA and HHV-6 DNA (23–26). However, the upfront and assay validation costs associated with digital PCR make its potential role in routine diagnostics unclear, especially given that the real-time PCRs are already well established in the diagnostic laboratory.

Mass spec prospects

Taking a different approach, Abbott is combining sample preparation, broad PCR amplification, and electrospray ionization mass spectrometry (ESI-MS) of DNA amplicons in an automated platform to identify base composition based on molecular weight (27, 28). Previously, this technology was known as Ibis T5000 and Abbott PLEX-ID, but it was re-launched in 2015 as IRIDICA. With this new technology, bacteria, viruses, fungi, and protozoa can be screened against a library of more than 750,000 entries to perform high-resolution subtyping, identification of known virulence markers and antibiotic resistance genes, and identification of mixtures of microbes from a single sample. Currently, Abbott has focused the development and marketing of IRIDICA within bacterial sepsis, mycology and sterile fluid diagnostic bacteriology. A panel for plasma viruses with further panels for respiratory viruses and encephalitis viruses exist and these could potentially become competitive against established real-time PCR if the pricing is pitched right.

Looking to the future

There are a huge number of technologies that could provide real value to the clinical virology laboratory, especially when they are fully automated. Though I haven’t covered them all in this article, I believe it’s clear that there are some exciting developments ahead that will revolutionize the way we diagnose viral diseases. Personally, I believe the future lies in further automation and miniaturization of enzyme immunoassays and nucleic acid amplification, providing either qualitative or quantitative results.

Perhaps the hospital laboratory will evolve to support a syndrome-based POCT service by developing community-based near-patient-testing projects so that family physicians and community nurses have access to the technology. For decades, physicians have sent samples away to the laboratory for testing and awaited results and interpretation. Future physicians are likely to interpret patient test results on their own and follow set clinical pathway protocols, thereby hopefully decreasing the overall cost of hospital stay. In fact, this progression is already happening in places where molecular POCTs have been installed in hospital wards. Once these technologies are widespread and interconnected, a truly integrated healthcare system is possible.

Where would hospital-based laboratory testing fit? It makes sense that laboratory medicine expertise and technologies be directed towards the management of more complex and multiple pathogen infections, novel pathogen discovery in idiopathic syndromes, providing high-volume tests at low cost, delivering automated algorithm-based interpretation of difficult sets of results and widespread real-time use of genome sequencing for infection control purposes. There is no reason why POCT and laboratory diagnostics could not co-exist in a mutually beneficial way. It’s already happening in some hospitals and my prediction is that it will only continue to become increasingly standard in the months and years ahead.

Malur Sudhanva is consultant virologist at South London Specialist Virology Centre, Viapath, King’s College Hospital NHS Foundation Trust, London, UK and honorary consultant virologist at Rare and Imported Pathogens Laboratory, Public Health England Microbiological Services, Porton Down, UK.