473 
Research using biochip analysers capable of detecting several molecules simultaneously reveals that they perform well against existing testing methods and are cost-effective, especially to help treat patients with rheumatoid arthritis
Anne Marie Dupuy
MD, PhD
Jean Paul Cristol
MD, PhD
Biochemistry laboratory
Lapeyronie Hospital
Montpellier
France
Montpellier University Hospital Centre (CHRU) is in the southern French region of Languedoc Roussillon, nationally recognised as an important source of scientific research. Together with the Provence-Alpes-Côte d’Azur region, it is the largest regional centre for scientific research in terms of financial support (6% of the budget is allocated to research) and the second largest hospital-university group for clinical trials, accounting for 16% of the country’s research in life sciences.[1,2] Montpellier CHRU has 30% of the total capacity of the department (county) with 2,700 beds and is the largest employer in Languedoc Roussillon.
The hospital is considered a centre of excellence in clinical research because of the large number of national research organisations (CNRS and INSERM) and university and hospital laboratories involved in research – along with the general environment of scientific and technical quality.
The department of biochemistry, led by Professor Jean Paul Cristol, is part of the hospital’s pathological biology group and has a budget of €1,620,000. Cutting-edge technology is an area of interest within the group, and priority is given to the development of new tools for the diagnosis, prognosis and general care of patients.
We are particularly interested in the recently developed multiplex analysers capable of detecting several molecules simultaneously in a single reaction. Multiplexing is based on innovative technologies in R&D use for several years now in the form of ‘biochips’ or ‘arrays’ developed for analysis of the transcriptome (RNA). The measurement of multiplex biomarkers is known to lead to significant savings in costs and the amount of biological samples. The further development to ‘proteomic’ biomarkers using sensitive immunodetection techniques provides the possibility of establishing the real profiles of the proteins expressed and the ability to find the links with diagnosis, treatment and prognosis.
With this in mind, we have tested biochips in different situations. The first was under emergency conditions where the chip was designed to detect biomarkers for cardiac or renal insufficiency/failure. Our emergency services tested cardiac chips – the ‘Triage System’ from Inverness Medical Innovations – designed to help the rapid diagnosis and assessment of heart failure and in risk stratification in acute coronary syndromes and heart failure. No sample preparation is needed and blood can be taken and analysed directly at the point of care. The arrays measure the total troponin 1, creatine kinase MB (CK-MB), myoglobin and B-type natriuretic peptide (BNP). Following a specific protocol, staff used the biochips either in the ambulance or at the bedside, when called to a suspected heart attack. The idea was that on-the-spot testing would save time and give a better diagnosis and prognosis than ECG.
Rapid predictor
On arrival at the hospital, another confirmatory test was carried out in the biochemistry department. In fact, we found the results at the incident scene and on admission were exactly the same, so earlier sampling and testing were not under our conditions of use, improving patient care.
In collaboration with our colleagues in the intensive care unit, we tested the NGAL (neutrophil gelatinase-associated lipocalin) test biochip. Like the cardiac chip, the test can be performed at point of care: anticoagulated whole blood or plasma is applied directly to the chip and a result obtained within 15 minutes. We found that NGAL was a more rapid predictor of acute renal failure than serum creatinine in patients with sepsis, and therefore this test does appear to be of value in improving patient care.
Since 2004, we have used the semi-automated Evidence Investigator from Randox in our laboratory. We have tested three different protein biochip array technologies (PBAT) from this system in extensive research projects to justify their future inclusion in routine use.
The first array we tested is not a medical array but one designed for use by the food industry, the Anti-Microbial Array. This simultaneously detects 12 sulphonamide antimicrobials which may be used in apiculture to treat bacterial foulbrood diseases. Sulphonamide antimicrobials are generally banned or strictly regulated, because residues found in honey can cause skin rash, hives, pruritus and anaphylaxis. The Anti-Microbial Array is designed to enable rapid and sensitive analysis of honey samples before processing. We tested honey from three different beehives and found no trace of sulphonamides.
The second array we have tested is for drugs of abuse. This directly tests urine samples for the presence of the banned substances frequently screened for in sport or the army. The main substances of interest are cocaine, opiates and cannabinoid derivatives. A team, led by Professor V. Gardet and Dr A. Servonnet, compared the results obtained using the biochips against the reference technique of gas chromatography-mass spectrometry (GC-MS) and found good correlation.
Use in rheumatoid arthritis
The final panel – and the one that we have used the most extensively and tested in the most different situations – is the cytokine array that simultaneously screens for 12 cytokines. The array includes the pro-inflammatory cytokines: IL-1α, IL-1β, IL-2, IL-6, IL-8, IFNγ, TNFα, the anti-inflammatory cytokines: IL-4 and IL-10, and growth factors: monocyte chemoattractant protein (MCP)-1, epidermal growth factor (EGF) and vascular endothelium growth factor (VGEF).
This panel is of great interest to us in the area of inflammatory diseases, in particular for rheumatoid arthritis (RA) where the array serves a double purpose. It can be used for the detection of the imbalance between Th1 and Th2 which is characteristic of autoimmune diseases such as RA, where pro-inflammatory cytokines are in excess and there is a deficit of inflammatory cytokines. We have also shown the array can be used to predict which patients will respond adequately to TNFα blocking agents (inhibitors), popular but expensive first-line biologicals used in RA.
Being able to correctly predict which patients would benefit from this type of therapy not only has economic advantages, but also saves patients the time and disappointment associated with treatment failure. We published our first paper of the predictive use of the cytokine array in 2008.[3] We showed that high serum levels of MCP-1 and EGF were associated with a response to etanercept and also that an increase in the combined parameters of EGF and C-reactive protein gave an accurate and sensitive prediction of etanercept responders at three months. This is exciting because, for the first time, we have a tool for predicting response in patients with long-standing or very active RA and obviously it opens the way for further research using other parameters and biological therapeutics.
Cytokines are also important in other disorders where inflammation is considered to play a central role in their pathology, such as chronic renal insufficiency and osteoporosis. These are newer disease models in which we are testing the cytokine panel.
Narcolepsy research
We are also using the panel to screen patients with narcolepsy, recently begun to be considered as an autoimmune disease. In collaboration with Professor Dauvilliers at the Reference Centre for Sleep Disorders, we have been studying the cerebrospinal fluid (CSF) and serum of patients for evidence of an imbalance between pro- and anti-inflammatory cytokines. We hope the arrays will be sensitive enough to enable serum markers to be detected, which will avoid the invasive techniques needed to obtain CSF, where the markers are generally found. Currently, we have data but no definitive results.
One important thing to remember when using biochips is that each chip arrays for 12 different cytokines. This means that in a study large enough to be statistically relevant, the amount of data generated will soon be enormous and requires expert help from a statistician to interpret it correctly.
We are extremely positive about the value biochips can bring to research and eventually to routine patient care. Multiplexing offers enormous advantages over classical enzyme-linked immunosorbent assay (ELISA) techniques in terms of technician time and reduced biological samples. Although the initial outlay is high and arrays are not cheap, the cost becomes equivalent once more than four cytokines are needed to be screened.
One issue is the lack of flexibility in the available arrays. This is a complex problem related to current biomarker research and approval procedures, patenting and the technology required to attach protein markers to chips. All of these will doubtless be overcome in the next few years, but today there are limitations to the use of this exciting novel technology which we need to be aware of. The value of biochips today lies in using them to identify the most relevant marker(s) from within a group of major biomarkers for the eventual rapid, routine screening of target populations.
References
1. Observatoire des sciences et des techniques (www.obs-ost.fr). Rapport 2004: Indicateurs de sciences et de technologies.
2. Audit conducted by consultants Ernst & Young.
3. Fabre S et al. Clin Exp Immunol 2008;153:188-95.
