The Peripheral Biomarkers special series, published in Alzheimer’s Research & Therapy, tackles the current issues surrounding the search to identify blood biomarkers for AD and discusses biomarker approaches that have already produced data that may help to advance the field.

Douglas Galasko, Director of the UCSD Shiley-Marcos Alzheimer’s disease Research Center, USA.

In this accompanying podcast, Alzheimer’s Research & Therapy In-house Editor Kathryn Smith talks to Series Editor Douglas Galasko about how attitudes towards biomarker studies have changed over the years, current exciting avenues of investigation and key questions in the field. Galasko also shares his thoughts on the future directions of biomarker research in AD.

Galasko is a neurologist at the University California, San Diego, USA, Director of the UCSD Shiley-Marcos Alzheimer’s disease Research Center and is also co-Editor-in-Chief of Alzheimer’s Research & Therapy. As Series Editor for the ‘Peripheral Biomarkers’ special series, Galasko brings together a collection of articles addressing a variety of topics, including; the potential of plasma amyloid beta (Aβ) as a biomarker, the use of biomarkers to evaluate the risk of incident AD, and novel approaches to biomarker discovery.

“In the 1980s we had a reasonable idea of the pathology of Alzheimer’s disease and yet this was not being used at all to make the diagnosis. So I was interested in trying to see whether the key pathological features of Alzheimer’s, namely plaques and tangles, would be reflected in something we could measure as a biomarker.”
Douglas Galasko, University of California, San Diego

Highlights of the series include a review by John Trojanowski and colleagues from the University of Pennsylvania School of Medicine, USA on plasma measures of Aβ, summarising a wealth of data on Aβ as a diagnostic and a predictive biomarker, as well as addressing the demographic, clinical, genetic and technical issues surrounding Aβ levels and measurement

In addition to covering the benefits and drawbacks of individual biomarkers, the series also takes a look at multiplex biomarkers in a review by Ralph Martins from Edith Cowan University, Australia and colleagues. The challenges faced by researchers in developing a global multiplex panel of biomarkers for AD are summarised and moreover potential solutions are outlined with the aim of advancing the field by reducing variability across studies.

Efforts to make progress in biomarker discovery have also led to studies into the body’s response to abnormal pathological proteins. Thomas Kodadek and Lorraine Fuhrmann Clark from the Scripps Research Institute, USA discuss key questions around the identification and screening of autoantibodies produced in response to various pathological features of AD.

Upcoming series content includes a review by Sudha Seshadri and Galit Weinstein from Boston University, USA discussing the candidates for potential use as circulating biomarkers that predict incident AD.

“There is great interest in biomarkers in general, and there have been many demonstrations of their value in approaches to diagnosis and therapy,” remarked Galasko in regards to the importance of peripheral biomarkers in AD. More on these advances and current issues faced in peripheral biomarker research for AD can be found in the ‘Peripheral Biomarkers’ special series.

The complete list of series articles:

Peripheral Biomarkers

What was your motivation for performing the study?
This research is mainly driven by the appreciation that miRNA is a rather stable component not only in serum but even more so in blood cells. Our study is conducted as part of a general endeavor towards the development of miRNA blood born biomarkers. Within this framework we set out to identify disease specific miRNA signatures.

What did you find most surprising about your results and what can  we learn from this?
Most intriguing is the disease specificity of the miRNA pattern found in blood cells of patients. At this point of our research we are only at the beginning of a biological understanding of the miRNA pattern identified. Similarities in some elements of the miRNA pattern between Alzheimer’s disease patients and those with other neurological conditions, possibly reflect similar biological processes in the cells harboring the miRNAs. Knowing the identities of the miRNAs present in blood biomarkers will allow us to gain insight into aspects of the immunological response, at least for those cell types that contribute most of the miRNA signatures.

How close to the clinic do you believe miRNA blood biomarkers for Alzheimer’s disease are?
Our results make us confident that miRNA signatures can likely play a role in the future diagnosis of Alzheimer’s disease. Prior to clinical studies it is however important to clarify the biological role of the identified miRNAs and to determine to what extent the miRNA pattern reflects different stages of the disease.

Do you believe it will ever be possible to sufficiently partition an Alzheimer’s disease signature from other cohorts to make miRNA blood biomarkers the sole basis of diagnosis?
To answer this question we need to deepen our understanding of the biology of the blood born miRNA signature. Independent of the progress in our understanding of the function of miRNAs in blood cells, it is however likely that miRNA signatures will mostly play a role as companion diagnostics.

What’s next for your research?
Our primary near-term goal will be to address the question about the diagnostic potential of miRNAs as early biomarkers.  Given the situation of Alzheimer’s as a progressive disease, it appears of foremost importance to identify those miRNA that indicate the disease in the early stages.

For more on the burden of Alzheimer’s disease and efforts underway to better manage and treat the disease, see our interview with Executive Director of Alzheimer’s Disease International, Marc Wortmann.

How did this research come about, and why is there a need for a tool to improve communication and palliative care in ICUs?

Communication and support for patients and their families are of central concern in healthcare. However, evidence shows that these are often poorly addressed, especially at times of rapid changes in health status and during clinical uncertainty. The Intensive Care Unit is a setting in which patients are profoundly ill and where their circumstances can change rapidly. Previous research has highlighted concerns in relation to support for patients and families, lack of adequate control of patient’s symptoms, inconsistent or lack of communication and attention to patients and families’ individual wishes and to dignity, respect and peace in the Intensive Care Unit. This led us to believe that there was a need to improve communication and palliative care in Intensive Care Units, and to develop a tool to try and achieve this.

What are the challenges associated with communication within an ICU setting?

The Intensive Care Unit is a highly challenging environment for patients and families. Patients are normally profoundly ill making communication very difficult. The nature of that illness and its severity means that it is often difficult or impossible to communicate with them directly – many are for example unconscious and in the process of being ventilated. Their families are entering a very unfamiliar place and are often in shock about what has happened and struggle to cope with this while supporting the patient as best they can. Although the central goal of the Intensive Care Unit is to preserve or extend life, the nature of illness or trauma means that sometimes this is not possible. Staff strive to actively treat the person’s illness, and support them, and seek to extend life. However, the severity of the illnesses that individuals face means that sometimes this is not possible. Things can change quickly, and there is a lot of information for families to take in.

What is the PACE tool and how does it work?

PACE stands for psychosocial assessment and communication evaluation. The goal of developing this tool was to improve assessment and communication for all patients in the ICU, both those who may deteriorate and equally those who may recover. PACE comprises the training programme and an assessment, which is recorded in the clinical notes. The training programme involves collaborative work between the ICU and hospital palliative care team staff to look at ways to improve communication and how the PACE questions might serve as a prompt to improve social assessment and continue communication. The record within PACE asks for assessment of five aspects of care:

1) Family Details; such as children, relationships and guardianships.
2) Social Details; such as language, culture, but also financial concerns or even the need for transport and parking for the family.
3) Patient Preferences; including any previous wishes about who they would want to inform their treatment
4) Communication and information; how much the patient and the families are aware of the situation and who is key in a family  to be involved in communication. It also involves explaining what happens in the ICU and finally any other aspects which the patient, family or staff feel is important. The record then gives space for a continuing log of any communication updates. This is so when the clinical  team change during the 24 hour period, everyone is kept up to date.
5) Any other issues; a list providing useful resources as a prompt for staff to receive additional support if needed. PACE is completed by the key worker for the patient, usually a nurse within 24 hours of admission.

In your study you have tested PACE in an ICU setting. How could PACE be implemented on a wider scale and in other settings?

We tested PACE in ICU because it is where there is a lot of uncertainty for patients and families. But similar needs for good communication and psychosocial care, and the presence of uncertainty, exist across the healthcare system. PACE helps the staff to get to know the patient and family better, and helps the family and patient feel that someone is interested in them and their wishes, as well as in the purely biomedical aspects of their condition. So we believe it could be helpful in other settings, especially in hospitals but this would need testing. In ICU there are higher levels of staffing than the rest of the hospital, so it may be more difficult to train staff to ask the questions correctly, and to listen well to families and patients’ answers and wishes. There may be a need to modify the training component of PACE or the questions it asks for it to best work in other settings.

How easy will it be to implement PACE into clinical practice and what will this involve?

The implementation of any tool is important as a tool is only as good as the implementation that supports it. We deliberately kept PACE short, so it couldn’t be a tick box exercise or a check list – we wanted the staff to think about what the assessment meant and to take the time to listen to patients and families. Our research in the implementation found that in general PACE was completed within 24 to 48 hours of admission (only about half of PACEs were completed within 24 hours and the rest within 48 hours). Most people found it easy to record the family history, the family condition and the social circumstances but more difficult to ask about prior preferences that the patient might have. This is where the training and support, which needs to be ongoing, is important.

What’s next for your research? Are their plans for a Phase III study of PACE?

We want to test PACE more widely and a Phase III study is exactly the right next step. We would like to try this in several centres – at the moment PACE has been tested in one hospital. We are about to plan an expansion of the work and seek potential partners. We want to see if there are benefits of PACE but also if there are any downsides to using it. So far, we are encouraged because it appears that PACE is not only feasible but it seems to improve communication, information, the families perception of the honesty of staff, and symptom control. The next test is to evaluate this further with comparative trials. Our development and preliminary evaluation was supported by the UK National Institutes of Health Research (NIHR), and we will need to approach them or a charity to support the next phase in the work.

You belong to the UK Juvenile Dermatomyositis Research Group (JDRG). What is the JDRG and what do you hope to achieve as a part of it?

The JDRG is a national group of paediatric rheumatologists and scientists who are interested in improving the understanding and clinical care  of patients with JDM, by sharing information on their patients, promoting research  and creating the largest cohort study of JDM patients in Europe.

What is Juvenile Dermatomyositis (JDM) and how rare is it?

Juvenile Dermatomyositis (JDM) is an autoimmune disease affecting approximately 3 children in every million per year. Muscles weakness and skin rash are the main symptoms of JDM and it affects every child differently with some children experiencing a mild form of the disease while others display a more severe disease progression.

What challenges exist in diagnosing and treating JDM?

As a rare disease, it is difficult to collect sufficient numbers of JDM patients to test new diagnostic criteria and plan randomised controlled trials (RCT) of new drugs. At present most treatments are based on expert opinion rather than trial data. On a more positive note, there was a recent large RCT of rituximab (anti CD20 monoclonal antibody) in adult and juvenile dermatomyositis, and international groups, such as the JDRG, are working on updating the diagnostic criteria originally published in 1975.

What did your study set out to investigate?

We wanted to find out if myeloid related peptide (MRP) would be an accurate biomarker for disease activity in JDM, and if so, what might be the mechanism that underlies this observation.

What most excited you about  your findings?

We were very pleased to see that serum MRP did correlate with other measures of disease activity in JDM. MRP is heat stable and easy measured by ELISA which makes it an ideal candidate for a biomarker in clinical practice.

What impact would a specific biomarker have on the day-to-day management of this condition?

One of the major challenges of JDM is trying to distinguish muscle weakness caused by active inflammation from muscle damage caused by corticosteroids or deconditioning. Having a specific biomarker for disease activity, would allows us to limit treatment escalation to those patients with real disease flares, and in this way, minimise iatrogenic harm from our treatments.

Your study defines a pathway by which macrophages ‘crosstalk’ and perpetuate inflammtatory myocytis. Do you think this is specific to JDM?

No, this is unlikely to be specific to JDM. Interestingly, muscle inflammation is also a feature of the muscular dystrophies and it is possible that MRP contributes to pathology in these disorders.

What’s next for your research?

We are carrying out a wider study to identify prognostic factors in JDM, by examining serological markers of inflammation (such as MRP), muscle biopsy results and autoantibody status, to identify those patients at risk of poor long term outcomes. In this way, we hope to target therapy to the correct patients, saving aggressive treatments for those patients who really need it.

What led to your interest in ecological research and in particular airborne and satellite mapping?

I’ve been an ecologist for 20 years, coming into my field originally from the ground-based perspective interested in biodiversity, invasive species, and ecosystem function. However, throughout this period, I have also recognized and quantitatively documented the often extreme bias associated with ground-based measurements and observations. To address these issues, I have combined field, airborne and satellite approaches in ways that hopefully allow us to not only ’scale up‘ to larger geographic areas, but also to obtain measurements that simply could never be acquired in the field alone. The airborne technology that my team and I have developed actually can deliver quantitative information on vegetation structure and chemistry that cannot be obtained on the ground, at any scale, even in a field plot.

How have carbon stocks been traditionally measured?

Traditionally, carbon stocks have been estimated in the field using tape measures and wood samples. We measure the diameters and heights of the trees, take wood samples for density analysis, and then apply allometric equations to estimate the carbon stock of a forest. It is a crude process even with the very best field techniques and allometric equations. This approach, while in extremely widespread use worldwide, has inherent errors of sometimes 20 percent or more, even under the best conditions.

What is LiDAR and how does it work?

LiDAR stands for Light Detection and Ranging. It is a technology that uses emitted laser pulses to measure the time, and thus the distance, between objects. We can adjust the wavelength of the lasers so that they penetrate the forest canopy from above, allowing us to measure the height of the vegetation as well as the vertical partitioning of the foliage within the canopy.

Why is it useful to integrate LiDAR estimates with field inventory plots?

Certain types of airborne LiDAR, such as the Carnegie Airborne Observatory system, allow us to very precisely measure the height of the forest canopy and also the vertical partitioning of the foliage and other elements like branch architecture. We have developed scientific methods using these canopy structural measurements as stand-ins for the types of measurements we traditionally make in field inventory plots. From these, we can easily scale up from plots to larger landscapes and regional levels.

Why did you choose to test your method on carbon stocks in Panama?

We’ve been developing and testing our approach in many regions of the world, including Hawaii, California, Madagascar, South Africa, Costa Rica and throughout the Amazon basin. However, Panama offered two new and unique opportunities to apply our latest, improved carbon mapping methodology. First, Panama presented us with a chance to do an entire nation. That alone is a powerful statement in the context of carbon and resource policy development as well as climate change mitigation activities. Second, Panama hosts the Smithsonian Tropical Research Institute’s field plot network, which offered us a chance to test the accuracy of our airborne and satellite mapping methods.

How do you think your findings could impact policy for climate change mitigation programs?

The Panama project is the first in the world to provide proper estimates of uncertainty for measurements taken in high-resolution, nationwide carbon mapping. No other project has done so. The errors are quite low – 10-20 percent in any given hectare (2.5 acres) of land, which is the critical scale and threshold of uncertainty needed for policy-makers to ‘make deals’ based on carbon stock measurements for the use of forests to sequester carbon from our atmosphere -  thereby reducing our climate change footprint worldwide.  This is the critical scientific contribution to their process. Panama proved that it can be done.

What other applications do you predict high-fidelity carbon mapping will have in the future?

As ecologists, we know that vegetation carbon stock is an expression of a huge range of processes ranging from evolution to land use change. By mapping and monitoring the carbon locked up in and/or released from vegetation, we can understand the biogeography of plant communities, the role of natural disturbances in mediating the flow of carbon, nutrients and water in ecosystems, the role of humans in managing ecosystems, the impact of policy decisions on natural resources, and much more. Carbon is a keys component of all ecosystems and their functions.

What’s next for your research?

I am currently mapping the biomass and the biodiversity of the Western Amazon. I plan to reveal it over the next few years in ultra-high fidelity. I hope it will inspire everyone, from presidents to preschoolers, to think about 3.5 billion years of evolution leading to the most amazing ecoregion on our planet.

This study is part of the Prevención con Dieta Mediterránea (PREDIMED) study, what is this and how did it come about?

The PREDIMED trial is a multicenter, randomised primary cardiovascular prevention trial using Mediterranean diet. During 2002, following a call for proposals by the Spanish government, different Spanish investigators working in nutrition from different perspectives joined forces to start a research network. We applied together for a grant to start a large randomised trial to test the effectiveness of a Mediterranean diet on the ‘primary’ prevention of cardiovascular disease

What led you to look at the association between alcohol consumption and depression?

There is a hypothesis that some risk factors and mediator mechanisms are shared by depression and coronary heart disease. This hypothesis prompted our research on common dietary exposures for both depression and coronary heart disease. It is well known that moderate alcohol consumption is related to lower risk of coronary heart disease.

You found that consuming low to moderate amounts of wine, in the range of two to seven drinks per week, was associated with a lower risk of depression. Why do you think this might be?

The mechanistic explanations on how alcohol in low to moderate amounts may prevent unipolar depression are largely unknown. Moderate ethanol intake is known to increase serum high-density lipoprotein cholesterol, and to exert beneficial effects on endothelial function and inflammation. These mechanisms are likely to be  involved not only in coronary heart disease but also in depression. We thought that the higher concentration of active polyphenols in wine may account for higher protection by wine than by other alcoholic beverages. In addition, moderate alcohol drinking has a GABAergic effect, acting on GABAA receptors, that may prevent or counteract the effects of depression on this system.

Previous studies have found a direct association between alcohol consumption and symptoms of depression. Why do you think your results differ?

The direct association between alcohol intake and risk of depression in previous studies was only seen at higher levels of ethanol intake, i.e. only with heavy drinking.  But most of our participants who were drinkers, only consumed low to moderate amounts of alcohol. In fact, when we assessed the risk among those who drank more than 40 g/day of alcohol, we found that they were at higher risk, but this association was not statistically significant, probably due to the small number of heavy drinkers in our cohort. The differential characteristics of our study population need to be taken into account, because participants in the PREDIMED trial were an older, traditional Spanish Mediterranean population, who consumed chiefly wine, and mainly in a context of socialisation with family or friends. Subjects with problematic use of alcohol were explicitly excluded from this trial.

Why do you think the effect on the risk of depression was more pronounced with wine consumption in comparison to other alcoholic drinks?

The stronger inverse association for wine may be due in part to the lower average quantity of ethanol consumed among drinkers, but it can specially be strengthened by the presence of considerable amounts of polyphenols in wine. In this line of thought, wine consumption has also previously been related with stronger beneficial cardiovascular effects. In addition, preferentially wine drinkers are known to be at lower risk of becoming heavy drinkers.

Is it possible that there are alternative factors which may explain your results?

A potential alternative explanation might be that wine drinkers are healthier than non-wine drinkers in other aspects. But this is not likely to be a very likely explanation in our study, because we took into account the potential confounding effects of these factors, including lifestyle variables, health indicators and dietary intervention, and adjusted for them. Another potential alternative explanation might be related to different patterns of consumption among wine-drinkers than in other drinkers. In fact, preference for wine drinking has been associated with a less concentrated pattern of consumption, and this can account for a healthier pattern of consumption that avoids binge drinking with its potential harmful neurotoxic effects.

What are the implications of these findings and what further research is needed?

The implications are that moderate alcohol drinking, specifically in the form of wine, may represent a means of preventing unipolar depression; but further prospective studies and trials are needed to confirm our findings.

How did you become interested in research into emergency and critical care?

Following my surgical medical training, I chose to join the Critical Care and Trauma Center of the Osaka Prefectural Hospital Organization. Approximately 24,000 trauma deaths occur in Japan annually, making trauma the fifth leading cause of death, and among young adults, trauma is the leading cause of death. I became interested in trauma care, because I hope to decrease the number of trauma deaths, working closely with Assistant Professor Yasushi Nakamori of the Department of Emergency and Critical Care Medicine at Kansai Medical University. We have often experienced that pre-operative computer tomography (CT) has contributed to the development of effective intervention strategies for emergency bleeding control in cases of severe blunt trauma. We therefore became interested in investigating these benefits of pre-operative CT on mortality in severe blunt trauma patients requiring emergency bleeding control.

Can you elaborate on the controversy surrounding the use of pre-operative diagnostic CT in patients with severe blunt trauma?

Although x-ray and FAST (Focused Assessment with Sonography for Trauma) are readily accessible and widely used during primary evaluations, their sensitivity in recognizing a number of injuries is limited. CT has become the gold standard for definitive diagnostic imaging in severely injured patients. However, CT scanning has often required patient transfers, which is time-consuming and laborious. Trauma care guidelines suggest that CT scans, when indicated, are carried out as part of a secondary evaluation so as to not delay the resuscitation process. However, several approaches to installing CT scanners close to or inside emergency rooms have been presented and recent technical improvements have included increased scanning speeds.

What new insights does your study bring to light?

Although several studies have reported that CT in early trauma care improves outcomes, little was known about the impact of CT on mortality when pre-operative CT is performed for blunt trauma patients requiring emergency bleeding control, especially for patients at high risk of death and for hemodynamically unstable patients. We find that in these cases CT performed before emergency bleeding control is associated with significantly improved survival.

ATLS (Advanced Trauma Life Support) guidelines recommend CT scanning after emergency surgery in hemodynamically unstable patients, which goes against your findings. What mechanisms do you think underlie the improved survival you found in this group of patients?

Our study revealed that pre-operative CT might be associated with improved survival in patients at high risk of death (with a low score for probability of survival according to the Trauma Related Injury Severity Score) and in hemodynamically unstable patients. In patients with a high risk of death and more than one site of bleeding, survival rate in the CT group was significantly higher than that in the non-CT group. I think this is because CT helps in prioritizing the type of emergency bleeding control required in the first instance and is helpful in promptly tailoring subsequent treatment.

What are the challenges of conducting prospective interventional studies in emergency medicine?

In our study, the decision criterion for performance of CT and the imbalance in sample size may be the major biases. Because it is difficult to perform CT in hemodynamically unstable patients, the number of patients recruited in the CT group may often be lower than that recruited for the non-CT group. In interventional studies in emergency medicine, it is possible that these selection biases cannot be controlled entirely.

Are you planning any prospective studies on the use of diagnostic CT in blunt trauma patients?

We are planning a multicentre study to clarify in which populations with severe trauma, will CT have the most significant effect on patient outcome.

Technical advances have reduced CT scanning times. Do you think pre-operative CT scanning will ever replace x-ray imaging in the management of trauma in the emergency room?

In the trauma care environment of our study, the time needed for the CT scan including patient transfer time was about 20 minutes. Thus, I think that such CT scanning could not necessarily replace x-ray imaging in trauma care. In 2011, we implemented a new trauma workflow concept with a sliding CT scanner system with interventional radiology features (IVR-CT) in our emergency room. All life-saving procedures including damage control surgery and transcatheter arterial embolization (TAE) can be performed without relocating the patient. In the IVR-CT system, time to CT initiation was ten minutes, and times to the start of emergency bleeding control procedures were 45 minutes for surgery and 54 minutes for TAE – obviously shorter than those of our previous reports. We think that our IVR-CT system allowing faster diagnosis and definitive interventions will replace x-ray imaging in future trauma care.

How did this research come about and what was the main goal?

The incidence of cardiovascular disease and cancer, the two most frequent causes of death worldwide among non-communicable diseases, is constantly increasing. New biomarkers are needed for improved risk prediction in the general population, in order to identify the individuals who would benefit the most from clinical screening and risk factor control. Inflammation plays a central role in both cardiovascular disease and cancer and inflammatory biomarkers, in particular high sensitivity C-reactive protein (hsCRP), have been correlated with disease incidence and mortality. However, measurement of hsCRP only adds a mild improvement to the already established risk scores based on traditional risk factors such as smoking, diabetes, blood pressure and plasma lipids.

Procalcitonin is a pro-inflammatory and immunosuppressive mediator that is involved in the pathogenesis of bacterial infections and sepsis. Elevated procalcitonin levels in patients already diagnosed with acute coronary syndrome and cancer were shown to be correlated with increased mortality.  The aim of our study was to elucidate whether procalcitonin may be used as a biomarker to predict total and cause-specific mortality in apparently healthy individuals. Additionally, we performed a comparative analysis in order to assess whether measurement of procalcitonin offers additional prognostic information compared to hsCRP, as an already established inflammatory biomarker.

What is already know about procalcitonin?

Procalcitonin is the pro-hormone of calcitonin, which is involved in calcium metabolism.  Procalcitonin is mainly produced in the neuroendocrine C cells of the thyroid gland and in the K cells of the lung, but human adipose tissue is also able to secrete procalcitonin under the influence of certain stimuli. Procalcitonin concentrations are low in healthy individuals but increase dramatically in systemic inflammatory response syndrome (SIRS) associated with sepsis, trauma, surgery, burns and pancreatitis. In healthy individuals procalcitonin is associated with the components of metabolic syndrome, such as obesity and insulin resistance. We have recently shown that several of the traditional cardiovascular risk factors are correlated with plasma procalcitonin levels and that procalcitonin is associated with the incidence of coronary events in apparently healthy individuals (J Intern Med. 2012, 272(5):484-91).

What are the main findings of your study?

We studied the relationship between procalcitonin, hsCRP and the incidence of all-cause, cardiovascular and cancer mortality in 3322 middle-aged individuals with no previous history of cardiovascular disease or cancer, recruited from the general population. We demonstrate that procalcitonin is strongly associated with the risk of all-cause and cancer mortality and with the incidence of colon cancer in men, independently of smoking, diabetes, hypertension, Body Mass Index, plasma lipids, renal function and hsCRP. In contrast, hsCRP was independently associated with cardiovascular mortality, but not with cancer mortality in men. After adjustment for potential confounding risk factors and each other, neither biomarker was independently correlated with incident mortality in women.

How does this research add to our knowledge about procalcitonin?

We demonstrate for the first time an independent relationship between procalcitonin and mortality risk in healthy individuals, in line with previous findings in cardiovascular disease, cancer and sepsis patients. Importantly, this relationship was independent of hsCRP, suggesting that the two biomarkers might be involved in different pathogenic inflammatory pathways and provide complementary prognostic information. Although significant, the relationship between procalcitonin and the incidence of colon cancer should be interpreted with caution and needs to be confirmed in larger cohorts, due to the low number of incident colon cancer cases within our population.

What do you think could be the underlying mechanisms to explain the association of procalcitonin with all-cause and cancer mortality?

As opposed to CRP, procalcitonin has inherent inflammatory and immunosuppressive properties and might directly participate in the pathogenesis of cardiovascular disease and cancer. Procalcitonin stimulates the secretion of the pro-inflammatory cytokines TNFα and IL-6, which have been shown to play important roles in atherosclerosis and to link chronic inflammation and tumor development. In turn, TNFα is a strong stimulator of procalcitonin production, creating potential pro-inflammatory positive feedback loops. Procalcitonin has also been shown to be able to inhibit human monocyte and neutrophil mobility and might thus impair beneficial anti-tumor immune responses.

How do you think your findings could influence clinical practice in the future?

Our study reveals for the first time important independent correlations between the levels of the inflammatory mediator procalcitonin in middle-aged apparently healthy men and the future risk of cancer death. Unfortunately, we could not establish a clear cut-off value for procalcitonin as a cancer mortality predictor with an adequate sensitivity/specificity balance for direct use in clinical practice. However, the association between procalcitonin and the incidence of colon cancer in men might have important clinical implications for future cancer screening protocols, but needs to be further confirmed in studies adequately powered to answer this type of question.

Interestingly, we demonstrate clear differences between plasma levels of the inflammatory biomarkers procalcitonin and hsCRP with regard to the risk of death in healthy individuals. Procalcitonin was independently correlated with the incidence of cancer death and hsCRP with cardiovascular death in men, but neither biomarker was correlated with the risk of death in women. These important gender-specific and biomarker-specific differences should be taken into account in the design and interpretation of future clinical studies on cardiovascular disease and cancer.

What’s next for your research?

We will continue to focus our attention on the involvement of immune and inflammatory mechanisms and mediators in human disease, in particular cardiovascular disease and cancer. Although several inflammatory biomarkers have been proposed to correlate with morbidity and mortality in healthy individuals, none so far have shown sufficient specificity and sensitivity to enter clinical practice. A more individually-tailored approach is required in order to accurately predict incident disease in healthy individuals, combining genetic analysis, traditional risk factors and new biomarkers. The relationship between procalcitonin and the incidence of colon cancer in men is intriguing and needs to be confirmed in prospective case-control studies that particularly target this diagnosis. Additionally, it will be interesting to determine whether procalcitonin is linked to death due to a particular type of cancer or whether it reflects unspecific susceptibility to cancer mortality.

How did the Assemblathon come about?

The Assemblathon developed as an offshoot of the Genome 10K project, which is aiming to sequence the genomes of 10,000 vertebrates. There is almost no point in attempting to sequence so many genomes if we are not reasonably sure that we can accurately assemble those genomes from the initial sequencing read data. As there are so many different tools out there to perform de novo assembly, it seemed to make sense to try to benchmark some of them.

Why is it important to assess genome assembly?

There are many areas of genomics where a researcher can find a plethora of bioinformatics tools that all try to solve the same problem. However, different software tools often produce very different answers. Even the same tool can generate very different answers if you explore all of its parameters and configuration options. It is not always obvious how the results from different tools stack up against each other, and it is not always obvious as to which tools we should trust the most (if any).

If you wanted to know who made the best chili in your local area, then you could organize a ‘chili cook off’. As well as deciding an overall winner, you could also award prizes for specific categories of chili (best vegetarian chili, best low-fat recipe etc.). What we can do for chili we can also do for genome assemblers.

Contests like the Assemblathon can help reveal the differences in how different genome assemblers perform, and can also pinpoint the specific areas where one program might outperform another. However, just as tasting chili can be a very subjective experience, there can be similar issues when evaluating genome assemblers. One of the objectives for the Assemblathon was to try to get a handle on what ‘best’ means in the context of genome assembly.

What was the rationale for using a different approach to Assemblathon 1, where you used synthetic data?

It helps to put a jigsaw together when you have the picture on the box to help you! Most of the time when people perform genome assembly, they don’t have that picture. So it can be hard to know whether you even have all of the genome present (let alone whether it is accurately put together). In Assemblathon 1, we wanted to know what the answer was going to be before people started assembling data, so we used an artificial genome that was created in a way that tried to preserve some properties of a real genome.

However, for Assemblathon 2 there was a lot of interest in working with real world data. The genome assembly community wants to be solving problems that can actually help others with their research.

Why did you pick the three species that you did (budgie, boa constrictor and cichlid fish)?

A major factor was simply the availability of suitable sequencing data. But it also seemed a good idea to use some fairly diverse species, the genomes of which might pose different types of challenges for genome assemblers (different repeat content, heterozygosity etc.).

What were the main findings of Assemblathon 2 and were you surprised by the results?

To paraphrase Abraham Lincoln: you can get an assembler to perform well across all metrics in some species, across some metrics in all species, but you can’t get an assembler to perform well across all metrics in all species.

Personally speaking, I was expecting to see variation in performance between assemblers but I thought that some of the bigger ‘brand name’ assemblers might have shown more consistency when assessed across species.

The results from each of the teams on each of the genomes was very mixed, but would you say there were some entries that were better overall?

We were very diplomatic with how we described the outcome of Assemblathon 1, where some assemblers performed consistently better than others. But in Assemblathon 2, there was so much variation it seemed hard to say that any single team should be declared a winner. At best we can say that in a given species, and when considering a specific genome assembly metric, that there were winners i.e. some assemblers will give you longer contigs for the fish genome, some others might capture more of the genes in the snake genome, and others will give better coverage of the bird genome. These may all be different assemblers.

What is the appeal of making your work freely available in arXiv, and the open peer-review that GigaScience carries out?

Personally, I am a strong advocate that results from tax-payer funded research should become publicly available a.s.a.p. I think we can demonstrably show – by the volume of blog posts that have now discussed the Assemblathon 2 project – that the resulting conversation about genome assembly has been a useful one. I really hope that more journals adopt open peer-review and encourage the use of pre-print servers.

Do you think there will be an Assemblathon 3? If so, what would you do differently next time?

I’m in the process of writing a blog post on this very topic (to be posted here). There are many things that could, and perhaps should, be done differently. For starters, if the community embraces the FASTG format, it would make sense to use this as the format of choice for Assemblathon 3. Perhaps more importantly, we should find a different lead author!

For more on Assemblathon 2 and the unusual ‘meta’ peer review process that accompanied publication of the results in GigaScience please see Biome’s interview with the Editor-in-Chief of GigaScience, Laurie Goodman here.

What led to your interest into research in genomic medicine?

GG I worked as a fly quantitative geneticist for 15 years. When the revolution in human genomics came in the late 2000s, I decided the time was right to pursue a new challenge and to see whether my perspective might translate to human biology. I started with some basic genomic anthropology and did not expect to have a clinical focus so quickly, but when I moved to Georgia Tech in 2009, their partnership with Emory on development of a Predictive Health Institute provided the perfect opportunity to establish a genomic medicine program.

AB I started my research career as the Human Genome Project was just picking up steam. It seemed obvious ten years ago, and even more obvious now, that the high-throughput measurement of molecules, whether proteins, RNA, and now especially DNA, was going to influence how we medically diagnose and treat individuals.

What are the main goals of the research at the Center for Health Discovery and Well Being?

GG The Center for Health Discovery and Well Being (CHDWB) was initially conceived by Dr Ken Brigham to support the evaluation of innovative medical care models that place the emphasis on wellness and patient self-knowledge. It combines unusually deep annual clinical evaluations with a health partner model, where participants develop personalized health action plans focusing on those aspects of their health that need attention to prevent the onset of chronic illness. There are almost 700 healthy adults involved and this cohort provides multiple opportunities for long-term profiling of changes in clinical parameters with age.

Personalized medicine has largely focused on disease. However, your study focuses on sequencing healthy individuals. Can you tell us why this is so important?

GG Our core belief is that if individuals can be encouraged and empowered to be involved in their own medical care, they may adopt more healthy behavior patterns before they start aging. We are thus interested in using whole genome sequencing, in combination with annual CHDWB clinical profiling, and functional genomics profiling (so called transcriptomics and metabolomics), to generate individual profiles of health. Our pilot data suggests that people are hungry for this information, and our next step is to evaluate how they deal with it, whether it does encourage lifestyle changes, and eventually whether it improves quality of life. The idea is that common genotypes in the genome can be used to classify which aspects of a person’s health may have the highest risk for disease, and to combine this with their clinical profiles to make informed decisions.

AB As far as medicine has advanced in recent years, we still haven’t found any way to avoid dying! And today, you just can’t die without having something wrong with you or even being labelled with a disease, at least on your death certificate. So not to be too morbid or flippant, but we will all be patients someday. If there is a chance that genome sequencing can yield reasonably accurate predictions at increasingly lower costs, and if those predictions suggest low-cost interventions for health, then I would argue that the use of DNA sequencing in healthy individuals should be studied.

We use tools like colonoscopy, mammography, and prostate-specific antigen in preventative medicine, each of which is well known to have pluses and minuses in costs, reliability, and accuracy, and so it seems obvious that DNA sequencing will play a role in prevention eventually. I do fully acknowledge that it is still a hypothesis (a testable hypothesis!) as to whether patients are more likely to follow age-old advice about weight loss and smoking, with genomes in hand.

What was most surprising or interesting when you started looking at the results of combining sequencing data and clinical records?

AB It was surprising that predictions for future disease can be made in an automated way across individuals, and it was surprising that these predictions could yield preventative measures.

GG Given that we know that the genotypes discovered so far only explain a small fraction of risk for common diseases (such as asthma or diabetes), it is remarkable that there was something like a two to one ratio of concordance between genetic classification and clinical profile to discordance. This will only get better over the next five to ten years.

What resources will be needed in practice to translate complex personalized sequencing data into clinically relevant actions that promote health and well-being?

GG Money, desire, and proof of principle. We’re a long way from a clear funding model for this in a healthcare environment that tends to focus on curing illness rather than preventing its onset. The cost of genotyping is no longer prohibitive, but good clinical profiling is several thousand dollars a year, and the two need to be performed hand-in-hand. Next we need  participants to stay engaged and not lose interest as the initial benefits and novelty wear off. We hope that genetic information will be a part of the solution, but of course we need to be able to provide it in an informative manner, not just as a report delivered to their iPad. Both of these things are of course more likely if we can generate clear examples where this approach makes a difference: perhaps people who recognize that their hypertension is probably behavioral rather than genetic or who pay more attention to diabetes risk or discover the cause of their propensity to gain weight. We’re definitely interested in partnerships to help make this model a reality.

AB I think current physicians might be concerned about the amount of effort and diligence it will take to obtain and interpret genomes. I think we will eventually have tools to help them, in the same way we have visualization tools that represent radiological images in such a way to enable diagnosis. But it could be that specialists take on this role, the same way radiologists perform mammograms or gastroenterologists perform colonoscopies for preventative care.

How do you think your findings could affect clinical practice in the future? What do you think the benefits and challenges will be?

GG I’m under no illusions that this will happen overnight, but very optimistic that many people definitely want it to happen. I think we need to overcome the deterministic view of genetics and claims that we can use genetic profiles to predict disease, and instead measure success in terms of positive behavioral changes. Economic models suggest that it does not take a lot of improvement in general health to reduce absenteeism, the costs of prescription drugs, and to delay the onset of chronic disease, for this type of program to pay for itself. Since the benefit is more to the group than the individual (perhaps only a minority will actually respond) it is not clear who will pay. Then there are the scientific challenges: n=1 genetics is in its infancy and it is never easy to establish causality or even validate classifications and predictions. So patience is also going to be essential in the midst of the great excitement over genomic medicine.

AB It is important to test many forms of prevention, and this is just one step of many needed to improve preventive care using genomics. If future studies also demonstrate the utility of this kind of work, and show improvements in quality-adjusted life years (QALYs) with maximum accuracy and at minimum cost, then I can see that this type of genomics can reach clinical care.

What’s next for your research? Are there plans to expand your pilot study?

GG The next two challenges for us are to actually begin giving people their genomic profiles and evaluate whether it makes a difference, and to incorporate transcriptome and metabolome analyses into our genomic assessments. We’re planning on engaging tech-savvy students at Georgia Tech for the latter, and to work with a variety of physicians at Emory on the former. Then perhaps a third challenge will be to see if the whole approach can be used to help with long-term recovery from chronic illness. So much of genomics is focused on who will get disease, not enough on what to do to prevent it or to promote full recovery.

AB We have ideas to run future studies with increased numbers of individuals, as well as increasing the number of ‘omic’ technologies to be run.