Pharmacogenomics analyses the response of individual patients to a medicinal product, to optimize therapy, obtain maximum efficacy and minimize its side effects.
Pharmacogenomics is now increasingly accepted to encompass pharmacogenetics, which focuses only on heritable biomarkers. Unlike the latter, pharmacogenomics also includes the study of proteins and enzymes as biomarkers.
The proponents of pharmacogenomics believe it holds the key to personalized medicine, in which drugs are tailored to a patient’s unique genetic profile.
Roots of pharmacogenomics in Human Genome Project
Pharmacogenomics is among the first clinical applications of the ambitious Human Genome Project, which was completed in 2003. It has already begun making an impact on clinical medicine, and promises much more as new pharmacogenomic biomarkers are identified by increasingly versatile techniques such as single nucleotide polymorphisms (SNP), small nuclear (sn) RNA-mediation and others.
ADRs are research priority
Pharmacogenomic biomarkers are essentially DNA or RNA characteristics that measure normal biologic and pathogenic processes, as well as the pharmacologic response to drug intervention.
The highest priority of pharmacogenomic research is to identify biomarkers for adverse drug reactions (ADRs), which account for one-fifth of all readmissions to hospital and 4% of withdrawal of new medicines.
ADRs are among the leading causes of death, with as many as 100,000 deaths a year in the US.
Pharmacogenomic labelling of drugs
There already are over 120 drugs in the US which include pharmacogenomic biomarkers in their labels. In Europe, the number is smaller, about 35. One reason is that the European Medicines Agency, the pan-EU regulator, has limited authority in the area. This is because of the large number of drugs which have been approved by Member States (rather than the Agency), with updating of the drug label seen as their responsibility. However, “relabelling to include pharmacogenomic data does not seem to be a priority issue” for the regulatory agencies in individual EU Member States.
Pharmacogenomic labelling of drugs, nevertheless, has been standardized in both Europe and the US under three categories: ‘mandatory’ , ‘recommended’ and for ‘informative’ purposes. So far, mandatory pharmacogenomic labelling is required where clinical trials have established the basis for response. In the category of recommended use, there have been no clinical trials (so far).
Typically, biomarker labelling covers the following subjects: drug exposure and clinical response variability, risk of adverse events, genotype-specific dosing, polymorphic drug target and disposition genes.
Considerable attention has been given to biomarkers for a range of widely-used oncology products. Apart from trastuzumab, they include tamoxifen (for breast cancer therapy), irinotecan (metastatic colorectal cancer), panitumumab and cetuximab (colon cancer).
Pharmacogenomic research is also focused on a host of other drugs and drug classes: allopurinol (anti-inflammatories), flucloxacillin and amoxicillin clavulanate (anti-infectives), as well as statins and immunosuppressants.
Companion diagnostics: measuring response to therapy
Biomarkers have made it possible to sell so-called companion diagnostics alongside expensive drugs, so as to direct therapy to the most responsive patients. One of the most prominent examples is the HER-2 test, accompanying Herceptin (trastuzumab), used to fight metastatic gastric cancer. The drug costs €42,000 for a year’s treatment.
Companion diagnostics also enable identification of potential ADRs, for example tests for the HLA-B*5701 allele accompanying the anti-HIV drug abacavir and for HLA-B*1502 with the anti-epileptic carbamazepine. The latter poses a recently confirmed risk of Stevens-Johnson syndrome and toxic epidermal necrolysis (TEN) in Han Chinese and other Asians.
Drug development and relaunch
Pharmacogenomic biomarkers are becoming integrated tools in drug development, to assess pathways encoded by polymorphic genes and to identify the enzymes which lead to the formation of an active drug metabolite, before entering clinical trials.
One new application for pharmacogenomic data is the relaunch of drugs, which have been withdrawn because of adverse events. Novartis, for example, applied in 2009 to the European Medicines Agency to use Lumiracoxib in genetically selected populations. Lumiracoxib is a prostaglandin endoperoxide synthase 2 inhibitor. It was approved to treat osteoarthritis, but was withdrawn in 2005 because of cases of DILI (drug-induced liver injury). Although retrospective genetic analyses revealed that variants of the HLA-DQ allele could predict elevated transferase levels and identify patients susceptible to DILI, Novartis withdrew its application in 2011 due to its inability to provide additional data within the timeframe specified by the Agency.
Generic drugs and pharmacogenomics
Pharmacogenomics is proving to be a weapon against generic drug imports, especially from large, low-cost producers in countries like India.
In its first-ever Recommendation, the European Society of Pharmacogenomics and Theranostics (ESPT) has called for “a harmonized approach to an updatable drug labelling of generic versions for pharmacogenomic information, as is the case for the original drug.” The ESPT cites the case of Plavix (clopidogrel), used for dual antiplatelet therapy and once the world’s second bestselling drug. Pharmacogenomic information on Plavix, it states, “reveals that genetic polymorphisms of CYP enzymes … contribute to variation in the response of individual patients.” It concludes that pharmacogenomic labelling “should be extrapolated to all medications which are marketed as both branded and generic versions.”
Clinical labs have been late entrants
The role of the clinical laboratory in pharmacogenomics broadly encompasses the following components:
In spite of being a frontline player in the application of pharmacogenomics, the position of the clinical lab has been relatively muted and unrecognised.
In 2000, a feature article noted that the “clinical lab has rarely been discussed within the context of pharmacogenomics.” It however argued that, in the future, “clinical labs will be looked to for genetic test development and validation, and for high-throughput genotyping of patients in clinical trials and routine testing.” It urged “both the labs themselves and the industry as a whole” to take cognisance of the fact.
Different from classical genetic testing
Lab techniques for pharmacogenomics differ significantly from classical genetic testing through chromosome analysis. Although state-of-the-art microarrays can interrogate and evaluate vast masses of alleles, the interpretation of test results into clinically meaningful data is complex, sometimes bewilderingly so.
This is because a particular gene mutation does not always result in a predictable phenotypic effect. A host of non-genetic factors can also play an influential role. Included here are the age, gender and ethnicity of a patient; so too are interactions with other drugs he or she is taking, and above all, any impairment in areas such as liver or renal function.
Usable and actionable information on such diverse factors may only emerge after adequate throughput of clinical data and the establishment of correspondences between genotypic and phenotypic markers. The sole entity which can bring such a scale to being is the clinical laboratory.
Meanwhile, physicians too are overloaded by new diagnostic information which emerges by the day, and need guidance from laboratories on how best to interpret and use the information contained in tests.
Unadopted pharmacogenomic tests
One factor that could accelerate the need for more clinical lab involvement may be pharmacogenomic tests which have not been adopted, in spite of evidence that they work. The best examples here are the enzymes VKORC1 and CYP2C9 for the anti-coagulant warfarin, UGT1A1 for the anti-cancer drug Irinotecan. In such cases, there have been concerns that diagnostic test costs may overwhelm the healthcare system, without demonstrable benefit.
At the moment, it is principally academic groups which are addressing such challenges. The price paid here is an acceptance of the fact that pharmacogenomics will only be “adopted slowly as risk-benefit data demonstrate the value of testing.”
Lab tests and healthcare spending
There is a heated debate underway in the US about laboratory testing as a source of healthcare spending growth. In November 2013, researchers at Beth Israel Deaconess Medical Center (BIDMC) announced the results of a review of more than 1.6 million results from 46 of the 50 most common lab tests. They found that nearly one-third of all blood tests were unnecessary.”
Some experts believe that a solution to this problem might be to increase ‘useful’ tests by laboratories, above all those for pharmacogenomic biomarkers. Even if the growth of personalized medicine increases laboratory testing, they argue this will improve a physician’s ability to make highly targeted decisions about patient treatment. This, in turn, may well reduce overall healthcare spending.
Such perspectives were suggested by Ramy Arnaout, Assistant Professor of Pathology at Harvard Medical School, and lead author of the BIDMC study. He argues that “lab tests are inexpensive. Ordering one more test or one less test isn’t going to ‘bend the curve,’ even if we do it across the board. It’s everything that happens next – the downstream visits, the surgeries, the hospital stays – that matters to patients and to the economy and should matter to us.”
The emergence of drug-resistant bacteria is one of the biggest public health challenges today. In April 2014, the World Health Organization (WHO) warned that a “post-antibiotic era – in which common infections and minor injuries can kill” is far from “an apocalyptic fantasy” and has instead become “a very real possibility.” Two months later, the newly formed World Alliance Against Antibiotic Resistance (with over 700 members in 55 countries) launched an urgent appeal to label antibiotic resistance “a grave global threat, and asking that antibiotics be declared a cultural heritage deserving legal protection.”
The ‘horror’ of carbapenem resistance
The march of bacterial resistance seems unremitting, even as the pipeline of new antibiotics is drying up. What is especially worrying for public health authorities, however, is a new “horror”. This refers to the growth of bacterial strains resistant to carbapenem – the ‘last resort’ antibiotic for unresponsive patients.
In several countries, carbapenem does not work in more than half of the people with Klebsiella pneumoniae, a major cause of hospital-acquired respiratory tract infections.
During the month of June 2014 alone, the US saw its first case of carbapenem-resistant (CR) Pseudomonas aeruginosa. In Canada, routine testing of raw squid in Saskatoon revealed a bacterial strain resistant to carbapenem. The case was the first of its kind in a food store, and demonstrates an enhancement of exposure risk “from a relatively small slice of the public” – such as travellers to risk zones or those recently hospitalized – to a much larger sector, according to Joseph Rubin, an assistant professor at the University of Saskatchewan.
Resistant genes are the real challenge
The Canadian case also illustrates the real, long-term challenge facing microbiologists. The bacterium in question, Pseudomonas fluorescens, is not risky for people with healthy immune systems. However, it carries a gene to produce carbapenemase, the enzyme inducing resistance to carbapenem.
Indeed, the real problem is less the spread of antibiotic-resistant bacteria than of antibiotic-resistant genes. Bacteria swap “small bits of DNA that carry genes like those for carbapenem resistance” and then “quickly pass it on to other species through gene swapping.”
A ticking time bomb
The implications are stark, and pose Scylla and Charybdis scenarios for public health authorities.
For example, third generation cephalosporins have been shown to fail in treating gonorrhea in much of Europe, Australia and Japan. Given that over 1 million people are infected globally with gonorrhea, every day, this is clearly a ticking time bomb.
Nevertheless, the use of carbapenems seems ill-advised. As the Canadian Medical Association Journal noted in 2011, this is because “gonorrhea readily shares its antibiotic resistance genes”, and using carbapenems “would invariably result in increased resistance … in other microbes. This life-saving antimicrobial would then have been ‘wasted’ on non–life-threatening gonococcal infections.”
Genomics and molecular methods
The above challenges necessitate a sophisticated arsenal of tools in the microbiology lab. Fortunately, breakthroughs in biotechnology, especially DNA sequencing and genomics, have shown some paths to the future.
In 1995, clinical microbiology witnessed the launch of the genomic era when the first bacterial genome of
Haemophilus influenzae was sequenced. So far, “over 1,000 bacterial genomes and 3,000 viral genomes, including representatives of all significant human pathogens,” have been sequenced – leading in turn to “unprecedented advances in pathogen diagnosis and genotyping and in the detection of virulence and antibiotic resistance.”
Sophisticated genomic tools based on molecular array techniques have been key to this success. Backed by high-speed computing, automated platforms and bioinformatics software, they provide labs with a host of new capabilities.
Molecular methods principally address a major limitation with previous phenotypic methods: the long period (weeks or even months) required to isolate some slow-growing bacteria. Indeed, the effort on the Haemophilus influenzae genome in 1995 had taken over a year, and although the Sanger sequencing used for this had long been considered a ‘gold standard’ , it soon faced “inherent limitations in throughput, scalability, speed and resolution.”
NGS offers exponential leap in sequencing
One of the most promising new molecular techniques is next-generation sequencing (NGS), which is also referred to as second-generation sequencing or SGS.
In terms of its basic principle, NGS parallels the capillary electrophoresis (CE) used in Sanger sequencing, with DNA fragments identified by signals as they are resynthesized from a template strand. However, while CE is limited to one or a few DNA fragments, NGS uses massively parallel processes to cover millions, and sequence several human genomes in a single run within days. This allows for identification of DNA base pairs covering whole genomes, and compare genetic differences down to resolution of a single base pair.
The latest NGS systems can generate over 300 Gb per flow cell, discover SNPs and chromosomal rearrangements, conduct transcriptome analysis, generate expression profiles, detect splice variants and quantify protein-DNA interactions.
Predicting antibiotic resistance, rapidly
NGS has already been used as a frontline weapon to identify bacterial pathogens and conduct epidemiological typing to define transmission pathways and support outbreak investigations – including cholera in Haiti in 2010 and and E. coli O104:H4 in Germany in 2011.
The technique is also seen as a way to predict antibiotic resistance, complementing phenotypic tests with the high-speed investigation of anomalies in results. It also resolves some of the biggest hurdles confounding traditional techniques, which require isolating resistance from environmental samples by PCR amplification or the cloning of cultured bacteria.
Both techniques, however, ignore large reservoirs of potential antibiotic resistance. PCR is incapable of broad-spectrum screening and is generally limited to known resistance genes. A bigger problem is that several bacteria are simply not culturable. This is an especially major challenge. Given that most antibiotics are produced by environmental microorganisms, most antibiotic resistance genes are likely to also have emerged outside a clinical setting.
Along with the new field of metagenomics, NGS has been harnessed to directly address these limitations.
The promise of metagenomics
Metagenomics was developed in the late 1990s for the function-based analysis of mixed environmental DNA species – and the existence of genes or genetic variations causing resistance. Metagenomics was directly aimed to address limitations in culturing and PCR amplification.
In its early years, metagenomics was principally targeted at recovery of novel biomolecules from environmental samples. The emergence of NGS, however, opened up wholly new frontiers – above all, to allow a fraction of the DNA in the sample to be isolated and sequenced, without cloning.
Metagenomics has been used to identify a range of antibiotic resistance genes, including tetracycline, aminoglycosides, bleomycin and β-lactamase.
The fight against resistant bacteria is likely to intensify in the years to come. In May 2014, ‘Cell Biology’ published the results of “the largest metagenomic search for antibiotic resistance genes in the DNA sequences of microbial communities from around the globe.” The findings illustrate the scope of the challenge: “bacteria carrying those vexing genes turn up everywhere in nature that scientists look for them.”
Challenges ahead
NGS is no doubt going to play a major role in this process. In spite of its novelty, it has already proven to be user-friendly for clinical laboratories. Last year, researchers showed it to be capable of integrating whole genome sequencing into the routine, daily workflow of a laboratory and rapidly resolve complex bacterial populations. NGS succeeded in identifying all ‘mixed-sample’ organisms taken from primary isolation plates. These included strains with scarce reads, notwithstanding an extremely low depth of coverage across its genome.
However, there is still some way to go. Key issues include the need to establish a reference genome database to archive, access and exchange the massive amount of data which is still to be generated. Some experts have called for the database to be open and accessible to the global scientific community.
A related challenge is that of interdisciplinary skills and collaboration. The analysis of NGS data requires a variety of specialists, namely “clinical and biomedical informaticians, computational biologists, molecular pathologists, programmers, statisticians, biologists, as well as clinicians.” Given that laboratories or other single institutions are unlikely to have all these skills in-house, an open database would again be the best means to find collaborative solutions.
Limits to Moore’s Law
The greatest driver of NGS adoption will be cost. The Human Genome Project cost around USD 3 billion. NGS can reduce this to “a few thousand dollars”, and achieve it much faster.
So far, NGS seems to be one of a handful of technologies challenging Moore’s Law, which describes a long-term trend of computing power doubling every two years. Compared to $95.2 million in costs per genome estimated in autumn 2001, the figure has since fallen by over 25,000 times. The decline has however not been uniform. Typically, costs had been dropping at an average of 70% a year between 2001 and 2007. In 2008, however, they fell massively, to $342,500 per genome in October compared to $7.1 million the previous year. This was wholly due to the move from Sanger-based to NGS sequencing technologies.
Costs per genome were estimated to be just over $4,000 in January 2014.
NGS, of course, has a variety of uses. It is seen as a tool to revolutionize molecular biology, molecular epidemiology and predict the evolution of bacteria. However, the fight against the menace of antibiotic resistance is likely to be one of its highest profile uses in the years to come.
April | May 2025
Beukenlaan 137
5616 VD Eindhoven
The Netherlands
+31 85064 55 82
info@clinlabint.com
PanGlobal Media is not responsible for any error or omission that might occur in the electronic display of product or company data.