Glynn13a.pdf

Lab on a Chip View Article Online Published on 22 April 2013. Downloaded by Dublin City University on 29/07/2013 10:07:25. CRITICAL REVIEW Cite this: Lab Chip, 2013, 13, 2731 View Journal | View Issue CD4 counting technologies for HIV therapy monitoring in resource-poor settings – state-of-the-art and emerging microtechnologies3 Macdara T. Glynn, David J. Kinahan and Jens Ducrée* Modern advancements in pharmaceuticals have provided individuals who have been infected with the human immunodeficiency virus (HIV) with the possibility of significantly extending their survival rates. When administered sufficiently soon after infection, antiretroviral therapy (ART) allows medical practitioners to control onset of the symptoms of the associated acquired immune deficiency syndrome (AIDS). Active monitoring of the immune system in both HIV patients and individuals who are regarded as ‘‘at-risk’’ is critical in the decision making process for when to start a patient on ART. A reliable and common method for such monitoring is to observe any decline in the number of CD4 expressing T-helper cells in the blood of a patient. However, the technology, expertise, infrastructure and costs to carry out such a diagnostic cannot be handled by medical services in resource-poor regions where HIV is endemic. Addressing this shortfall, commercialized point-of-care (POC) CD4 cell count systems are now available in Received 15th February 2013, Accepted 22nd April 2013 such regions. A number of newer devices will also soon be on the market, some the result of recent maturing of charity-funded initiatives. Many of the current and imminent devices are enabled by microfluidic solutions, and this review will critically survey and analyze these POC technologies for CD4 DOI: 10.1039/c3lc50213a counting, both on-market and near-to-market deployment. Additionally, promising technologies under www.rsc.org/loc development that may usher in a new generation of devices will be presented. CD4+ T cell counts as an indicator of AIDS Biomedical Diagnostic Institute, National Centre for Sensor Research, School of Physical Sciences, Dublin City University, Dublin, Ireland. E-mail: macdara.glynn@dcu.ie; jens.ducree@dcu.ie 3 Electronic supplementary information (ESI) available. See DOI: 10.1039/ c3lc50213a Acquired immunodeficiency syndrome (AIDS) was linked to infection with the human immunodeficiency virus (HIV) in the first half of the 1980s,1–3 and since then over 60 million infections have occurred, leading to over 30 million deaths.4 Macdara Glynn earned his PhD from the National University of Ireland Galway (NUIG) examining DNA damage response pathways of human fibroblasts in response to platinum-based chemotherapeutics, and followed this with a postdoctoral fellowship at the Centre for Chromosome Biology where he led projects on epigenetic inheritance mechanisms of the human centromere. Macdara moved to Stokes Bio Ltd. as the Macdara Glynn Senior Cellular Biologist, where he ran the cell biology facilities and was a member of a multidisciplinary team designing a massively high-throughput droplet based qPCR system. Macdara is currently a postdoctoral researcher at the Microfluidic Platforms Group at Dublin City University. David Kinahan received his PhD from Stokes Institute, University of Limerick in 2008 for work on fluorescent melting curve analysis of DNA in microchannels. He then joined Stokes Bio Ltd (a spin-out from Stokes Institute that developed high through-put qPCR instrumentation and which in 2010 was acquired by Life Technologies) as a Senior Engineer. The primary focus of his work was instrumentation David Kinahan and firmware development. In late 2010 David was promoted to Engineering Manager; leading a team of 10 engineers which included 4 post-doctoral researchers. In January 2012 David joined the Microfluidic Platforms Group at DCU as a post-doctoral researcher. This journal is ß The Royal Society of Chemistry 2013 Lab Chip, 2013, 13, 2731–2748 | 2731 View Article Online Published on 22 April 2013. Downloaded by Dublin City University on 29/07/2013 10:07:25. Critical Review Development and access to anti-retroviral therapies (ART) increased the survivability of infection, and in recent years led to significant reductions of mortality in patients when treatment regimens were maintained.5,6 However, of the 7000 new HIV infections a day reported in 2010, roughly 97% occurred in low and middle income countries.7 This is consistent with overall epidemiological statistics showing that low- to mid-income countries consistently maintain the largest viral pool of HIV, with the Sub-Saharan African region being the most heavily affected (69% of the 34 million global cases in 2011).8 A significant barrier to reducing HIV penetrance in low- to mid-income countries is diagnosis and monitoring. A reliable diagnostic of HIV (and the subsequent emergence of AIDS) is a measurement of the number of CD4+ expressing T-helper cells (Th-cells) in the patients’ blood–the primary host cell of HIV infection. As the disease progresses, the number of these cells decline in the blood. Physicians hence use a CD4 count to determine if a patient should begin ART, and also to monitor the recovery of a patient undergoing therapy. Another diagnostic of AIDS is the measurement of HIV viral load in a patient, but this review will focus only on CD4 cell count. There are two useful measurements to report the levels of CD4 cells in a patient: absolute CD4 count, and %CD4. Absolute CD4 count delivers a value indicating the number of CD4 cells per microliter of blood and would normally be in the range of 500–1200 cells ml21.9 %CD4 measurements indicate the percentage of all leukocytes that are CD4 cells. As the total leukocyte and CD4 count is higher in young children than in older children or adults, %CD4 is the preferred method of diagnosis for patients ,5 years of age as the percentage of CD4 cells does not fluctuate as much as the absolute CD4 count. There are guidelines for use of absolute CD4 counts in infants in resource-limited settings, however these are recommended only when %CD4 is not available.10 In patients older than 5 years, either absolute CD4 or %CD4 can be used, but absolute CD4 is preferred.11 Prof. Jens Ducrée is the Associate Professor for ‘‘Microsystems’’ and Principal Investigator for ‘‘Microfluidic Platforms’’ in the Biomedical Diagnostics Institute (BDI) at the Dublin City University (DCU). Before moving to Ireland, Prof. Ducrée initiated and led the lab-on-a-chip group and the lab-on-a-chip polymer microfabrication foundry service at HSG-IMIT and IMTEK in Freiburg (Germany). His major Jens Ducrée fields of interest are microfluidic lab-on-a-chip technologies for system integration and automation of life-science applications. Prof. Ducrée has published .150 peerreviewed publications, two books, 7 book chapters, more than 20 invited talks, and is (co-)inventor on 45 patent filings. 2732 | Lab Chip, 2013, 13, 2731–2748 Lab on a Chip Table 1 Recommended thresholds for ART initiation ,24 Months 24–59 Months 5 years - Adult Absolute CD4 (CD4 per ml blood) All % CD4 All (% of total leukocytes) ,750 ,350 ,25 N/A Since 2009, it has been recommended by the World Health Organisation (WHO) that patients presenting with CD4 counts below 350 cells ml21 begin ART regardless of clinical symptoms,12 while patients with counts less than 200 cells ml21 are considered to be presenting with AIDS.13 Table 1 summarises the recommended cut-offs for ART initiation for patients across different age groups. Challenges in delivering CD4 counts for all There are currently a number of technological, economic and societal challenges to deliver cheap and accurate CD4 counting to the remote regions of low- and middle-income countries and to ensure that the long-term commitment to monitoring is maintained both by the patient and the clinic. The ‘‘goldstandard’’ for CD4 counting is a rapid 3- or 4-colour test carried out on a flow cytometer in a moderately equipped medical clinic or hospital. These instruments deliver a highly accurate measurement but are immobile and too expensive to be useful as a core instrument in resource-poor settings or remote regions. Additionally, such instruments require skilled operators to run/maintain/repair and also to interpret the data generated. As the considerable majority of current and future AIDS patients reside in developing countries, there is a striking need for a reliable Point-of-Care (POC) diagnostic for CD4 cell counting that is sufficiently portable, cheap, easy-to-use, robust and accurate that can be deployed in sufficient quantities to allow minimally-trained operators to monitor individuals in their home-villages rather than patients traveling to distant clinics. As well as tackling the economic challenges of accessible diagnostics in developing countries, a CD4 POC diagnostic may address a societal issue in which up to 80% of patients drop out of treatment coverage between diagnosis and treatment, many of these drop-outs are associated with travel distance to the clinic and high death rates.14,15 Following introduction of a POC CD4 test at a clinic in Mozambique, an observational cohort study published by Jani et al. found that total loss-to-follow-up of patients dropped from 64% to 33%, and the median time from enrolment to ART initiation fell from 48 to 20 days.16 Another study compared the cost per life-year-saved (LYS) between standard flow cytometry and POC CD4 counting in Malawi. This study found that although the total cost of both strategies was similar, the cost per LYS was $148.30 for POC and $165.50 for flow cytometry.17 Together, these data identify POC CD4 testing as a promising strategy to lessen the impact of HIV infection in developing countries. The primary goal of a HIV POC is to bring the diagnostic capability to often unskilled operators in remote, lowinfrastructure locations. Motivated by its established capabil- This journal is ß The Royal Society of Chemistry 2013 View Article Online Lab on a Chip Critical Review Published on 22 April 2013. Downloaded by Dublin City University on 29/07/2013 10:07:25. Table 2 ASSURED criteria for POC diagnostic devices in resource poor regions CD4 POC recommendation18 ASSURED criteria Motivation A – Affordable N Instrument and per-test cost low enough to be affordable by those at most risk of infection S – Sensitive N High probability of a ‘treat’ result when CD4 is below 350 ml21 S – Specific N High probability of a ‘no treat’ result when CD4 is above 350 ml21 U – User-friendly N Minimal training required for non-medical personnel to operate the test N Minimal operator intervention to reduce required training and complexity N Use of autonomously filled sample cartridge/dip-stick is ideal N Simple assay protocol N Final metric (cells ml21 or %CD4 ratio) displayed/observed without operator calculation or interpretation required N Patient blood should be contained to protect the operator from blood-borne pathogens Portable Few steps Automated incubation Ideally finger-prick whole blood input R – Rapid and robust N Time from sample-to-answer is within the acceptable time of a patients standard visit to a clinic N Enable treatment at first visit N Reagents and instrument can be stored and operated without refrigeration, reducing power consumption and weight N Maintenance free/simple on-site repair N Sturdy construction to withstand off-road/dry/dusty/monsoon environments Self-contained ,30 min turnaround time Operate in adverse environmental conditions including high and low temperature ranges Heat stable reagents No cold storage E – Equipment-free N Outside of the instrument itself and the associated consumable if appropriate, minimal requirement for additional non-standard equipment (pipettes, centrifuge etc.) N Battery powered and adaptable to a number of recharge options (solar, AC, car alternator etc.) Battery or solar operated D – Delivered N Supply and pricing applied to ensure populations in most need of POC diagnostic devices have access to them N Increase economy of supply by supplying in bulk N Long shelf-life of consumable as some regions may be far from supply chains N Device tested and calibrated routinely without requirement for visiting engineer N Reagents used are QC tested at source Built in QC Ideally have electronic data collection and transmission ity of comprehensive process integration and automation of CD4+ cell enumeration on a low-cost, disposable cartridge interfacing with a low-footprint, widely autonomous instrument, various academic and corporate research teams pursue the development of a commercially viable microfluidic lab-ona-chip device as a HIV POC. Clinical and operational requirements for POC CD4 cell counting In order to be useful as a POC diagnostic in a resource limited environment, the WHO authored guidelines for tests under the acronym ASSURED (Table 2). An expert panel was convened by the WHO in June 2012 to make recommendations on development priorities for HIV diagnostics. This panel proposed general targets for implementation of the ASSURED criteria in POC diagnostics as the following: Affordable (,5 USD), Sensitive (99%); Specific (98%); User-friendly (requiring minimal operator training); Robust (no cold chain, no operator calibration, minimal routine maintenance), rapid (,1 h: same day result and care); Equipment-free (battery operated, no moving parts); Deliverable (commercially available and approved). The report further suggested characteristics specific to CD4 POC testing. Although not officially This journal is ß The Royal Society of Chemistry 2013 endorsed by the WHO at time of writing, these characteristics are included in Table 2 for reference.18 Venous and capillary sampling An important factor that must be considered for quality of data in POC devices is the method of sample extraction from patients. For blood based tests, the primary options are venepuncture and capillary extraction (generally via disposable lancet ‘‘finger-prick’’). For POC instruments, capillary extraction is generally favoured as: a) lower levels of operator expertise are required; b) higher levels of safety are afforded to the operator as needle-stick injury with potentially contaminated blood is reduced; and c) extraction is less invasive for the patient assuming that multiple finger-pricks are not required. Given that capillary and venous blood samples are not expected to differ significantly in terms of haematological counts,19 devices that aim to offer capillary blood as an input sample must hence study any bias of capillary versus venous blood on the system that arise from other factors such as operator or sampling error. A number of studies present data showing that capillary and venous blood can be used interchangeably for CD4+ counting,20,21 however conflicting reports show significant differences. Glencross et al. carried out an in depth study of the potential error introduced on a Lab Chip, 2013, 13, 2731–2748 | 2733 View Article Online Published on 22 April 2013. Downloaded by Dublin City University on 29/07/2013 10:07:25. Critical Review POC device when capillary blood is used instead of venous in South Africa.22 The Alere PimaTM device (see below) was tested a) in hospital with pipetted venous blood; b) in the same hospital with capillary blood; and c) ‘‘in the field’’ at lower- or higher-resourced primary health care clinics. For b–c, matching venous samples were also drawn from patients to directly compare capillary and venous sampling. When venous blood was used at the hospital, the study found tight negative bias with little variation around the bias on the device. Also, there was excellent correlation between a number of Pima instruments. However, when capillary blood was used on the same devices there was a larger negative bias with increased variation. Additionally, more than 10% of the capillary samples gave a ‘‘no read’’ result. When tested in the field, variance was increased further and patient samples were lost either due to capillary blood clotting or to ‘‘no read’’ results. In fact, during the field testing 10 out of 32 patients in the ,350 cells ml21 group had CD4+ counts reading above the clinical threshold when read using capillary blood on the device meaning these patients would have missed out on ART were the capillary result followed instead of the venous. Throughout these tests, the operation of the device itself was monitored using quality control cartridges, and confirmed the devices were operating with acceptable baseline reproducibility confirming sample handling and operator influence played a prime role in the variance shown in the results. These data are in general agreement with that from other studies;21,23,24 however, tests were shown to improve when operators were given intensive and on-going training for finger-prick blood extraction from the manufacturer, indicating that standardised extraction protocols are a key requirement for the rollout of a new POC device. Other than operator training, there are other factors that may influence increased variance when using capillary blood. The introduction of air bubbles during sample application to the disposable cartridge, or clotting of the sample can lead to issues with the internal microfluidics that may be alleviated using properly anti-coagulated venous blood introduced via pipette. Together, these studies demonstrate that capillary extraction can be a viable option for POC HIV diagnostics; but for each POC device, sufficient studies must be carried out to establish the tolerance of the result to variances in extraction protocol, positive or negative bias, and limits of agreement established compared to the ideal of a venous blood sample. Fundamental techniques in CD4+ POC devices A number of commercial instruments and academic strategies for CD4+ cell counting will be discussed throughout this review. Although each instrument and lab-based technology brings its own innovation to bear on the POC landscape, there are a number of underlying fundamental strategies and technical approaches that a number of the devices share. For sample input and processing, instruments will accept whole blood via pipetting or direct application to the test cartridge from a finger prick. The sample progresses along the microfluidic regions of the cartridge via capillary action or active pumping. Processing of the sample will generally involve either a) solid-phase immunocapture of the cells of 2734 | Lab Chip, 2013, 13, 2731–2748 Lab on a Chip interest (CD4+ and/or CD3+ cells); or b) solution-phase immuno-labelling of the target cells, sometimes followed by immobilisation or distribution in a detection location. For immunocapture-based instruments, the red blood cells and background white blood cells can be removed from the system by wash buffers while the target cells remain at the capture/ detection location. Alternatively, should the target cells be specifically labelled with fluorescent markers, they can be identified from the background material with minimal enrichment. Physical cell parameters are also utilised to immobilise or enrich target cells. For example, membranes that withhold target cells based on size while allowing the passage of smaller non-target cells are often used in technologies being developed in academic groups. Finally, detection and enumeration is often based on either optical or electrical strategies. The most common of the optical strategies is fluorescent based labelling and is a powerful but often expensive method of specifically imaging cells of interest. Electrical impedance can be used on a per-cell basis to physically count the number of cells passing a detector, or by measuring the difference in the impedance of a cell enumeration chamber as captured cells undergo lysis. Current technologies for CD4 counts in resource-poor regions As mentioned, flow cytometric CD4 cell counting is cumbersome and expensive. However, in addition to novel POC devices coming on to the market, it is worth also considering the ‘‘streamlined’’ flow cytometric systems that have been developed with characteristics more amenable to deployment in resource-poor countries. These systems were recently reviewed by Pattanapanyasat et al.25 The focus of this review will be on non-flow cytometric POC devices; however, a brief overview of some of the flow-based systems is included for comparison. Portable flow cytometry based systems Flow cytometry based instrumentation can be broadly divided into two classes: dual- and single-platform systems. With a dual platform, a flow cytometer (%CD4 value) is paired with a haematological analyser (absolute lymphocyte count) to generate an absolute CD4 Th-cell value. Due to laboratories using various analysers, these approaches lead to high variability in results between labs. Yet, these systems are quite common as core instruments in medical clinics. A single-platform instrument generates absolute CD4 or %CD4 values in a single tube using technology based on either precise sample volume measurement, or addition of numerically calibrated fluorescent microbeads to the sample. Single-platforms show greater reproducibility between laboratories than the dual-platforms, but reagent and machine costs can be higher. The more compact design of the single-platforms has led to them being adapted for portability and use in remote regions. This journal is ß The Royal Society of Chemistry 2013 View Article Online Published on 22 April 2013. Downloaded by Dublin City University on 29/07/2013 10:07:25. Lab on a Chip Critical Review Fig. 1 BD FACSCountTM flow cytometer. Image used with permission of Becton, Dickinson and Company. BD FACSCountTM – counting bead-based instrument A single-platform benchtop flow cytometer designed specifically for enumeration of absolute CD4 counts and %CD4 values is the BD FACSCount (Fig. 1).26 This is the only available such system based on micro-bead enumeration with a streamlined ‘‘no-lyse, no-wash’’ whole blood workflow. Reagents are supplied as ready-made monoclonal antibody (mAb) solutions with fluorophores precisely selected for the hardware within the instrument, and the protocols are predesigned with built-in software that automatically applies appropriate gates to the data. As with all bead-based systems, cell concentrations are calculated by adding a known number of fluorescent beads to a known volume of sample, allowing the volume-per-bead-detected to be calculated. Although more compact than a higher-level flow cytometer (i.e., it has no need for a computer, and results are printed on a sample report issuing from the instrument), BD FACSCount still requires relatively skilled operators and maintenance. It is not sufficiently portable to travel reliably to day clinics in remote areas, and QC using calibration beads must be run often, lowering the overall patient throughput. Partec CyFlow1 miniPOC – highly portable cytometer The German company Partec GmbH have released an impressively portable flow cytometer for POC HIV monitoring at primary health centres and remote areas – the CyFlow1 miniPOC (Fig. 2a). This hand-carried flow cytometry option offers a number of features that facilitate travel to remote areas, including light weight (,5 kg); ability to power from DC, rechargeable batteries or solar panels; fully integrated computational and printing facilities; and rapid result generation (40–70 s from sample loading). The manufactures claim this allows up to 250 CD4+ tests per day,27 and detects CD4+ counts in the range of 0–5000 cells ml21.28 The instrument reports absolute CD4+ counts as well as %CD4, making it useful for both adult and infant monitoring, as well as giving an option to deliver the sample data in a ‘‘routine mode’’ (in which a needle/dial graphic is displayed showing the CD4+ informa- This journal is ß The Royal Society of Chemistry 2013 Fig. 2 a) The Partec CyFlow1 miniPOC CD4 counting device. b) Images of the display in both expert mode (left), and routine mode (right). Images used with permission of Partec GmbH. tion), or an ‘‘expert mode’’ with a more familiar flow cytometric scatter-plot displaying the relevant gates applied (Fig. 2b). The active CD4+ detection reagents are dry stored in the sampling consumable and do not require cold storage. Although sample-to-answer times are short using this instrument, the sample preparation is performed off-instrument. These procedures are simple and do not require excessive training or medical knowledge, but involve a number of steps before the sampling device is loaded to the instrument and takes up to 15 min to complete. These steps do not include blood extraction which may have to be taken by a trained medical practitioner. As the input sample requires only 20 ml of EDTA treated whole blood however, it may be extracted via sterile lancets, allowing finger-prick blood extraction instead of venous extraction. As of the publication of the UNITAID 2012 HIV/AIDS Diagnostic Technology Landscape report, no peer reviewed and independent performance evaluations were yet released for the miniPOC instrument.28 Portable non-flow cytometry based systems Although instruments such as the Partec miniPOC mentioned above is an example of a portable CD4+ POC, generally flow cytometric systems do not lend themselves to being readily deployable for medical personnel travelling to remote areas. Despite the release of the PointCare NOWTM instrument in 2009 which brought remote POC CD4+ enumeration to the fore, this review will not focus on this instrument as it is a flow Lab Chip, 2013, 13, 2731–2748 | 2735 View Article Online Published on 22 April 2013. Downloaded by Dublin City University on 29/07/2013 10:07:25. Critical Review Fig. 3 Alere PimaTM Analyser: a) Representation of the dual CD3 and CD4 static image analysis technology. b) The Pima analyser with disposable chip. Image used with permission of Alere International. cytometry (albeit non-fluorescent based) instrument. We mention it here to place some context on the time-scale of true POC technologies being deployed from 2009. For the following instruments, integration of advances in both microfluidics and miniaturised signal detection has been a primary enabler for development. This combination resulted in instruments that are highly specific to a particular role (CD4+ cell counting in this case) rather than depending on an expensive multi-assay capable item of capital equipment. Herein lies a principle strength of microfluidic diagnostic solutions: dedicated and affordable instruments that can proceed from design to manufacture in a relatively short space of time. Alere PimaTM Analyser – static fluorescent image analysis Perhaps the first non-flow based dedicated POC CD4+ analyser suitable for deployment in remote regions is the PimaTM Analyser, manufactured by Alere Inc. and released in 2009 (Fig. 3b). The Pima instrument is a fixed-volume cytometer that uses static image analysis to generate a CD4+ count, but, notably, not a %CD4 readout, thus limiting its use in infant HIV monitoring.28 Like the Partec miniPOC, the Pima has the 2736 | Lab Chip, 2013, 13, 2731–2748 Lab on a Chip option to use a number of power sources including batteries and solar. Sample volume required is 25 ml of whole blood, and this can be taken from a finger-prick sterile lancet extraction, eliminating the requirement for an on-site phlebotomist. Once the lancet incision is made in the finger of the patient, the disposable Pima cartridge consumable is used to draw in the required volume of blood, where all the sample preparation and processing occurs without any further user intervention. Venous blood may also be applied to the cartridge directly. The workflow therefore does not require manual handling or processing of blood, and all occurs within the cartridge which isolates the sample from the instrument, thereby minimising chances of contamination of the instrument. With a maximum of 20 tests per day, throughput of the Pima may be considered low as compared to other instruments. The Pima is not fieldserviceable. However Alere maintain a stock of new instruments in a country for timely exchange on site.29 The technology of CD4 detection in the Pima is static image analysis based on LED illumination and multi-colour CCDbased detection of fluorescently labelled CD4 and CD3 cells in a whole blood sample (Fig. 3a).30 While a number of blood cell types may express either CD4 or CD3 antigens (i.e. monocytes and thymocytes, respectively), T-helper lymphocytes express both CD4 and CD3 simultaneously. Once the Pima cartridge is primed with the blood sample, monoclonal antibodies against both CD4 and CD3 are incubated with the blood, specifically labelling both epitopes with a distinct fluorescent marker on all cells that express either antigen. The cartridge is then loaded into the Pima Analyser, which excites the fluorophores with the on-board LEDs, and detects fluorescent emissions using the CCD camera (Fig. 3a). Proprietary image analysis integrated software then identifies and enumerates T-helper lymphocytes by identifying the cells that express both antigens. Results are expressed then as CD4+ cells ml21 blood.31 In a recently published study by Herbert et al. carried out at a HIV outpatient clinic in London UK, the Pima device was compared to laboratory CD4 testing in chronically and newly infected HIV patients. The study found that results from the Pima correlated strongly with those of the laboratory test (r = 0.93, p , 0.001), albeit with the Pima showing generally lower results overall. The sensitivity and specificity of the Pima were reported as 95% and 88% respectively.32 Additionally, performance of the Pima system has been evaluated and published independently in resource limited regions (Zimbabwe and Mozambique), and CD4+ counts were found to be comparable to the standard technology of the BD FACSCalibur. The accuracy of the instrument when operated by either a nurse or a lab technician was shown also to be comparable.21,23,29 The Mozambique study demonstrated that the reliability of the instrument when either finger-prick or venous blood was used as the input sample was also comparable.23 A more recent study using the Pima in an infectious disease clinic in Uganda recommended that the device is an attractive option for identifying patients eligible for ART, but suggested that negatively-biased CD4 counts at high absolute numbers may limit its use for the long-term immunological monitoring of ART progression in a patient.33 This journal is ß The Royal Society of Chemistry 2013 View Article Online Published on 22 April 2013. Downloaded by Dublin City University on 29/07/2013 10:07:25. Lab on a Chip Critical Review Fig. 4 The MBio CD4 instrument with associated multi-cartridge rack to allow parallel processing. Image used with permission of MBio Diagnostics Inc. MBio CD4 system – fluorescent imaging with parallel sample processing MBio Diagnostics Inc. (Boulder, CO, USA) has also developed a fluorescence-based CD4+ cell counter using an immuno-based dual-staining strategy (Fig. 4). As mentioned, many fluorescence-based technologies exist, but the expense and bulk of integrating the required optical detection system can limit its use in an easily portable POC device. MBio has overcome this limitation through proprietary technologies combining singleuse disposable cartridges with a simple and robust reader device. Finger-prick or venous whole blood is introduced directly to the MBio cartridge, which contains all required assay reagents. By incorporating lyophilized reagents within the cartridge, the system eliminates the storage and transportation cold chain requirement and facilitates a single-step workflow for sample processing. The lyophilized reagents include two fluorescently labelled antibodies, and the reader performs a two-colour image analysis of stained whole blood to provide an absolute count of CD3+ and CD4+ cells by excitation of both fluorophores. Use of this technology has allowed the MBio device to occupy a small POC-suitable footprint, while accessing the advantages of fluorescence based detection in a cell counting system. Although this detection strategy is similar to that of Alere’s Pima, the MBio CD4 System pursues a two-stage process in which the blood sample is processed in the disposable cartridge outside of the reader. Following sample processing, the cartridge is then placed into the reader to generate the final result. As the reader is not required for sample preparation, this allows the MBio CD4 system to carry out parallel processing of a number of cartridges at once through the use of an associated, multicartridge rack. Further innovations to make the device cost effective are the use of low-cost lasers similar to those found on DVD players, and optics technologies adapted from the mobile phone industry,34 and much of the device itself is of a plastic construction. Using a disposable cartridge with passive, capillary controlled fluidics (i.e. containing no pumps or This journal is ß The Royal Society of Chemistry 2013 Fig. 5 Daktari CD4 Counter. a) Representation of the microfluidic cell chromatography. b) Lysate impedance spectroscopy. c) Photograph of the Daktari disposable cartridge and associated reader device. Image used with permission of Daktari Diagnostics Inc. valves), throughput of the MBio CD4 system is reported as 60– 80 samples a day by a single operator using a single system.35 In collaboration with the University of California, San Diego, MBio Diagnostics performed field testing of an early prototype device in California, USA.36 The study used venous and capillary blood from HIV infected participants, and was compared to data obtained using a BD FACSCalibur flow cytometer.35 Results from the study showed only minimal downward bias in CD4 count on the MBio compared to the FACSCalibur when using venous blood, and even less bias when capillary samples were used. Some samples that were tested on the MBio system were mis-classified; however these samples were all within 100 cells ml21 of the 350 cells ml21 diagnostic threshold value. Daktari CD4 Counter – label-free microfluidic impedance detection Daktari Diagnostics Inc. (Boston, MA, USA–in collaboration with Continuum Advance Systems) has developed an instrument that goes without optics, lenses or cameras, hence the size and weight of the Daktari lends itself to extreme portability and a rugged construction (Fig. 5c). The lack of the expensive components listed above also suggests the price of the instrument could be minimal, and was reported as low as $800 in the UNITAID 2011 HIV/AIDS Diagnostic Technology Landscape Report.29 However, this estimated cost was elevated to $1000 in the 2012 report,28 and most recently to ,$5000 in the semi-annual update of the 2012 report.37 The increase in Lab Chip, 2013, 13, 2731–2748 | 2737 View Article Online Published on 22 April 2013. Downloaded by Dublin City University on 29/07/2013 10:07:25. Critical Review cost may indicate that the device may have been functionally upgraded to allow CD4 percentage measurements, and possibly HIV viral load to be measured on the same instrument. The device is powered by a rechargeable battery and requires no sample processing steps, allowing the instrument to be operated by minimally trained personnel. The CD4 enumeration technology is based on two applications: 1) microfluidic cell chromatography (Fig. 5a), and 2) lysate impedance spectroscopy (Fig. 5b). Together, these technologies enable the instrument to deliver a CD4+ cell count without the requirement for specific labelling of the cells.28 The disposable cartridge consumable that processes the patient sample is a microfluidic chip-like card that flows blood (and subsequent processing reagents) to appropriate locations in the cartridge by pressure generated via independent deformable blisters. A region on the cartridge features surface-immobilized anti-CD4 antibodies which capture CD4+ cells as the blood flows past the capture region. The chip is then flushed of all non-immobilised cells. This is the basis of the Microfluidic Cell Chromatography technology (Fig. 5a). The next step involves the Lysate Impedance Spectroscopy technology. Two reagents are released into the cell binding chamber resulting in cell lysis and release of the cellular components into the chamber. By measuring the electrical impedance of the resulting solution, Daktari software can estimate the CD4+ cell count in the original volume (Fig. 5b).38 The result generated on the Daktari is thus not a physical cell count, but an estimation of the original number present based on the concentration of cellular components released into the detection chamber. Data on the company website shows performance of a prototype instrument compared to that of a flow cytometer. Using the 350 CD4+ cells ml21 threshold, the prototype demonstrated a sensitivity of 0.90 and specificity of 0.97.39 Independent validation of the Daktari is not yet available at the time of writing; yet, the company plans to field-test the instrument in Kenya in the summer of 2013. Visitect1 – rapid point-of-care (RPOC) testing Recently launched at the AIDS 2012 conference in Washington USA, the Visitect is the result of collaboration between the Burnet Institute in Melbourne Australia and Omega Diagnostics Group in the UK. This device does not aim to count CD4+ cells directly, but measures the amount of CD4 protein in the sample. Assuming that the level of the CD4 surface marker present on individual Th-cells remains stable,40 monitoring the overall concentration of CD4 protein in a sample will give a diagnostic metric of Th cell concentration. The Visitect POC is a disposable device that accepts 40 ml of peripheral whole blood directly and uses lateral-flow technology to generate a colorimetric readout of CD4 protein concentration much akin to many commercial pregnancy tests (Fig. 6a). The single reagent buffer does not require cold storage. The outcome is a semi-quantitative ‘‘treat/no treat’’ answer that emerges after 40 min. An accompanying instrument may be purchased that allows traceability and storage of results (Fig. 6b), but this is optional and the tests may be interpreted by eye directly on the consumable POC device. 2738 | Lab Chip, 2013, 13, 2731–2748 Lab on a Chip Fig. 6 a) Interpretation of the result of a CD4 blood test on the Visitect diagnostic device. b) Image of the optional reading and data storage instrument. Images used with permission of Omega Diagnostics Ltd. When a blood sample is applied to the device, red blood cells and monocytes are retained on a pad at the site of sample application, while other cells (including the CD4+ cells) travel along the chip via lateral-flow. CD4+ cells are immobilised further up the device where they bind biotinylated anti-CD4 antibodies that have been rehydrated upon activation of the chip. A lysis buffer is added and the soluble CD4-protein/ antibody complex progresses along the device with its cytoplasmic domain now available for binding and immobilisation via a second anti-CD4 antibody located at the test line of the device. Finally, visualisation of the test result is facilitated by further addition of buffer which releases antibiotin molecules conjugated to gold particles. These particles bind to a) the immobile CD4/antibody complex at the test line, b) the reference line labelled ‘‘350’’ and, c) the control line at the top of the viewing area of the device. The intensity of the coloured band that appears at the reference line has been calibrated to represent the intensity of 350 CD4+ cells ml21 blood. If the test band is equal to or darker than the reference band, then a ‘‘no treat’’ result is given as this indicates that the patient has ¢350 CD4+ cells ml21 of blood. Conversely, if the reference band is darker than the test band, then a ‘‘treat’’ result is given (Fig. 6a). If the control band does not show up, or is broken, then the test must be repeated with a new device. As the result can be read by eye, the method is therefore instrument free if merely a ‘‘treat/no-treat’’ result is required. This journal is ß The Royal Society of Chemistry 2013 View Article Online Published on 22 April 2013. Downloaded by Dublin City University on 29/07/2013 10:07:25. Lab on a Chip Critical Review compacted mass of bead-bound cells is proportional to the initial concentration of the cells in the sample, and is simply observed by eye alongside a scale calibrated to indicate cell concentration similar to interpreting temperature using a thermometer. Anti-CD14 magnetic beads in the loading chamber also bind monocytes, and are immobilised by a magnetic collar. This minimises monocyte contamination that would otherwise lead to non-specific background signal, potentially resulting in CD4+ cell readout higher than the true count. Although this technology runs without detection hardware, it does require an associated mixer/spinner device that includes a lens to magnify the stack for ease of reading (Fig. 7b). At an estimated cost of $200 however (which is expected to be included in the cost of the purchase of a number of tests) this device should not be a financial overburden on a clinic.28 Additionally, a field version of the device will not require electricity to operate. The mixer/spinner device being developed is expected to perform 1 test in less than 10 min.34 Field trials suggest that the device is comparable to flow cytometry in medically relevant concentrations,43 and data pertaining to this is displayed on the company website by way of a correlation analysis chart. However, (at time of writing) these data have not yet been published in peer-reviewed journals. Fig. 7 a) The cell stacking disposable device from Zyomyx. The inset image shows how the result can be interpreted in a fashion similar to a thermometer. b) The electricity-free manually operated mixer/spinner accessory. Images used with permission of Zyomyx Inc. Data supplied on the manufacturer’s brochure shows the result of validation studies performed at the Alfred Hospital in Melbourne, Australia. Compared to a flow cytometric control, the Visitect correctly identified 97% of the ‘‘treat’’ patients, and 80% of the ‘‘no treat’’ patients. It is also planned to field test the device against flow cytometry and the Alere Pima. This study will be carried out at antenatal clinics in Sub-Saharan Africa, as well as a reference laboratory in South Africa. Zyomyx Inc. – electricity-independent diagnostics A potentially electronics-free POC CD4+ monitoring device is in development at the Californian company Zyomyx Inc. The test itself is comprised of a single disposable cartridge employing proprietary cell stacking technology, with an inbuilt reporting readout at the bottom of the cartridge that is viewed by eye (Fig. 7a). The system is somewhat akin to packed cell volume (PCV) haematocrit tests,41 albeit at microfluidic volumes. When blood is introduced to the device, heavy particles immunogenically bind specifically to the CD4+ cells, thereby increasing their individual density. Additionally, these particles are optically dense and so enable the visual readout at the completion of the protocol. The device is then mixed and spun such that only particles with the added density will permeate through a high-density medium in a wide capillary tube at the readout area of the device. Other cells remain above the high-density medium.42 The height, therefore, of the This journal is ß The Royal Society of Chemistry 2013 Assigning ASSURED criteria to current and near-market CD4+ counting POC devices As mentioned, the effectiveness of a rapid sample-to-answer POC device for deployment in a resource-poor setting should meet a number of criteria pertaining to this deployment. The WHO has issued such criteria – ASSURED – designed to reduce the cost of rapid POC testing of infectious diseases at the site of primary patient care rather than at a centralised laboratory.44,45 It is important to be aware that low scoring in the ASSURED criteria is not an indication of the quality of instrumentation or the data that is generated by a device. Rather, this criterion merely examines the option in terms of deployment to low-resource environments. The BD FACSCalibur, for example, may score quite low in the ASSURED examination due to its size, cost and required expertise; however, it is still regarded by most as the gold standard in experimental quality that the majority of other devices will be benchmarked against. Table 3 addresses the ASSURED criteria for each of the commercial devices described above. For comparison, the flow cytometry options have been included but only to demonstrate the effectiveness of using ASSURED to identify devices suitable for deployment to regions of limited income. Data was obtained from either published literature, company representatives, or estimated by the authors. In all cases, the manufacturers were contacted directly and asked to confirm the data. If no information is available or impossible to estimate, this is indicated in the table. The order in which the instruments are listed are as they are described above, and in no way indicates any score or ranking by the authors. Lab Chip, 2013, 13, 2731–2748 | 2739 View Article Online Critical Review Lab on a Chip Table 3 ASSURED criteria for current and near-commercialisation CD4+ counting POC devices Clinic suitable devices Published on 22 April 2013. Downloaded by Dublin City University on 29/07/2013 10:07:25. ASSURED criteria Affordable Instrument (US$) per test (US$) Sensitiveb Probability of a ‘‘treat’’ result when CD4 is below 350 ml21 Specificb Probability of a ‘‘no treat’’ result when CD4 is above 350 ml21 User-friendly # Stepsd Level of expertisee Rapid/robust Time to answer (min)f Storage Temp. (uC) Samples per dayg Equipment-free # Additional consumablesh # Additional instrumentsi Power Deliverable Independently verified (Y/N) Special deal for RPR Portable point-of-care devices FACSCalibur FACSCount miniPOC Pima MBio Daktari Visitect Zyomyx 75 000 3.00 30 000 3.50 9380 3.96 5500 6.00 ,5000 6.50 ,5000 8.00 1200a 5.00 200 ,8.00 1.00 1.00 — 0.95 — 0.9c 1.00 — 1.00 0.9 — 0.88 — 0.97c 0.83 — N/A 5 N/A 4 7 3 4 2 5 2 4 2 5–6 2 7 2 60 2–8 200–400 62–93 2–8 170 20–22 2–30 250 18–20 2–30 20 20 2–40 80 8 4–40 50 40 ¡40 120 ,10 2–40 40 3 3 AC 3 3 AC 6 0 AC/solar/ battery 3 1 AC/solar/ battery 5 0 AC/solar/ battery j 5 0 AC/solar/ battery 5 0 AC/battery (for reader) 5 0 None required Y Y N Y N N N N Agreements with WHO, Discounts for bulk ordering in place or expected for these portable point-of-care devices PEPFAR, Clinton Foundation etc. a Instrument optional. b ‘‘Treat/no treat’’ cut-off at time of writing is 350 CD4+ cells ml21 whole blood. c Data from a prototype device. Details available in electronic supplementary information S1. e Expertise: 1 – very low; 2 – low; 3 – medium; 4 – high; 5 – very high. f Including incubations. g Assuming 10 h day, 1 operator, 1 device. h Not including associated instrument consumable cartridge, but includes items for capillary puncture and wound dressing. i Including refrigerator if require. j Solar charger not offered as an option, but instrument is compatible with commercial solar chargers. d Future pipeline and market for POC CD4-based HIV diagnostics The challenge of delivering accurate and reliable HIV diagnostics based on CD4 cell enumeration in the absence of high-tech geographical and societal infrastructure has been taken up by a number of companies ranging from small startups to multinational corporations. Although the market could not be regarded as being saturated, the near future will see a choice of devices available for deployment in resource-limited regions; and a number of these devices (such as the Pima) are already available. In their annual and semi-annual Diagnostic Market Landscape reports, UNITAID projects the CD4 product pipeline for the next 2 years. The most recent version of this report predicts that the Visitect may be available as early as late 2012, but that 2013 will see the release of as many as 4 POC devices based on CD4 cell enumeration. These are the Daktari, MBio, Zyomyx and a new FACS based instrument from BD Biosciences called the BD FACSPresto.37 Added to the current options of the PointCare NOW, the miniPOC and the Pima – the release of these additional instruments to the market will significantly widen the availability of localised and accessible HIV monitoring in regions where it is currently endemic. If we consider the evidence that many of the patients 2740 | Lab Chip, 2013, 13, 2731–2748 who drop out of ART programs do so due to travel restrictions; if deployed extensively these devices should make an impact on the global response to control of HIV.14–16 Emerging technologies for CD4 counting in resource-poor settings Along with the commercial platforms described previously in this article, there has been a high level of activity within academia aimed at developing qualitative techniques to determine HIV disease stage and progression.42 Along with CD4+ lymphocyte count, these techniques include total lymphocyte count, HIV plasma viral load (PVL) and other surrogate markers.46,47 Initial work was targeted at validating and optimising commercial manual assays such as the Coulter Cytosphere assay for resource poor settings.48,49 However, recognising of the potential of microfluidics and bioMEMs50 in biomedical diagnostics, much effort has been made to adapt these strategies as a viable option for low-cost HIV diagnosis in resource-poor settings.51,52 This journal is ß The Royal Society of Chemistry 2013 View Article Online Lab on a Chip Published on 22 April 2013. Downloaded by Dublin City University on 29/07/2013 10:07:25. Membrane and micro-cavity isolation An extensive range of microfluidic sample preparation and detection strategies have been developed independently by various research groups. Rodriguez et al.53 developed a flow based system in which a 16.6 ml sample of whole blood (stained off chip with fluorescent antibodies) is passed over a track-etched membrane. The red blood cells (RBCs) pass through the membrane while the majority of larger white blood cells are trapped on its surface. Using the membrane as an optical plane, the white blood cells are imaged with three distinct wavelengths on a modified fluorescent microscope. Custom-built software is then used to identify cells expressing combinations of CD3, CD4 and CD8 markers. In an evolution of this design, Jokerst et al.54 used quantum dots as the fluorescent markers in an effort to reduce the cost of the optics associated with biological fluorophores. Alyassin et al.55 developed and applied custom analysis software to make quantitative, fluorescent based CD4+ counts from microfluidic platforms that immobilise CD4+ cells using antibody based surface chemistry, and also from platforms in which cells are captured on track-etched membranes. Hosokawa et al.56 have developed a micro-cavity array which contains 3 mm cavities separated by a 25 mm pitch. Under optimal conditions, leukocyte cells were individually trapped on the micro-cavities while remaining elements of a whole blood sample (such as erythrocytes and platelets) passed through the cavities to a waste reservoir. Fluorescently tagged cells were then imaged using a microscope and enumerated using data analysis software. Adhesion-based isolation and detection Adhesion-based detection has been recognised as a reliable cytometry solution suitable for point-of-care applications.57,58 Cheng et al.39,59 developed a flow-through system wherein 10 ml of whole blood was pumped through a microfluidic device (Fig. 8a). The microfluidic chip contained an enlarged chamber with biotinylated anti-CD4 antibodies coating its inner surface. Cells expressing the CD4+ cell surface epitope were immobilised to the inner walls of the chamber. Following a flush with a washing buffer, cells adhering to the inner surface of the chamber were viewed under a microscope and enumerated by investigators. This platform was further refined by addition of an upstream chamber for the depletion of CD14 expressing monocytes that, due to co-expression of CD4, can result in false positive readings.60 Thorslund et al.61 developed a microfluidic chip manufactured from PDMS which used capillary pumping to draw whole blood through a channel. During manufacture, the channel was initially coated with heparin to render the channel hydrophilic. A layer of avidin (which binds to heparin) was then immobilised on the surface. Finally, the channel was functionalised for CD4+ capture by binding biotinylated antiCD4 antibodies to the surface via interaction with the precoated avidin. Whole blood was loaded by capillary flow into the functionalised channel, and CD4+ cells were captured by the antibodies bound to the channel wall. Unbound materials were flushed from the channel by introduction of a washing buffer which was also controlled by capillary flow. Captured This journal is ß The Royal Society of Chemistry 2013 Critical Review CD4+ cells were specifically identified by staining using an anti-CD4+ fluorescent antibody, and total bound cells were visualized using a non-specific fluorescent nuclear stain. This platform was further developed62 by creating pillar-like structures in the micro-channel to increase the capillary forces which load the channels. Zhu et al.63 robotically printed specific-target antibody arrays (anti-CD3, -CD4 and -CD8; 150 and 300 mm diameter) on silanised or hydrogel-coated glass slides (Fig. 8b). These slides were incubated with RBC depleted whole blood. They found a strong correlation between leukocytes expressing specific epitopes (i.e. CD4) binding to a predictable location on the array corresponding to where the antibody to the target epitope had been printed. Using this location based gating strategy, CD4 : CD8 ratios were measured which showed close correlation with standard flow cytometry techniques. As fluorescent imaging is often a significant expense in the manufacture and maintenance of a cell counting platform, this strategy could obviate the need for fluorescent labelling, and hence potentially reduce the cost of cell-based immunological assays such as HIV testing. Investigations have also been made to increase the repeatability and reliability of CD4+ cell binding to chamber surfaces. Kim et al.64 developed a quartz nanopillar array (pillars of order 30–80 nm diameter) on which streptavidin was immobilised. They found that, due to increased surface area and interaction between nano-pillars and cellular microvilli, the percentage of captured CD4+ cells harvested from a mouse spleen was markedly increased over a control experiment conducted using bare glass. Gurkan et al.65 layered specific sections of a microfluidic device with both a thermally responsive polymer and neutravidin (Fig. 8c). A biotinylated anti-CD4 antibody was then bound to the surface via the neutravidin to enable CD4+ cell capture. CD4+ cells passing though the microchannel at 37 uC bound to the walls of the channel. A wash step removed all unbound components of whole blood. When the chip was then cooled to 32 uC, the thermally responsive polymer lost its ability to tether the antibody/cell complex to the surface of the chamber, and the bound cells were released from the channel walls. These cells were then washed through the device for collection. Paramagnetic beads have also been used to isolate CD4+ cells in microfluidic systems using a pseudo-adhesion approach. Furdui and Harrison66 developed a platform in which paramagnetic beads conjugated with anti-CD3+ were pumped through their system. At the upper and lower surfaces of their reaction chamber, permanent magnets attracted and subsequently bound the beads to the walls. A lymphocyte solution was then flowed through the reaction chamber and CD3+ cells were immobilized on the paramagnetic beads. The magnets were then used to physically move the CD3+ cells to a point on their platform where the cells of interest were removed using a pipette. Gao et al.67 adapted this approach whereby they isolated CD4+ cells in the reaction chamber (Fig. 8d). They then imaged the cells using a microscope and enumerated them manually. Li et al.68–70 and Ymeti et al.71 developed a platform whereby cells of interest are isolated immuno-magnetically. In this design, whole blood is incu- Lab Chip, 2013, 13, 2731–2748 | 2741 View Article Online Lab on a Chip Published on 22 April 2013. Downloaded by Dublin City University on 29/07/2013 10:07:25. Critical Review Fig. 8 Examples of microfluidic strategies for CD4+ isolation and enumeration based on cell adherence and visualisation. (a) Immobilised anti-CD4 antibodies adapted from59 with permission; (b) printed antibody arrays adapted with permission from;63 (c) CD4 capture and release adapted from65 with permission of The Royal Society of Chemistry; (d) paramagnetic bead binding adapted from67 with permission from Elsevier. 2742 | Lab Chip, 2013, 13, 2731–2748 This journal is ß The Royal Society of Chemistry 2013 View Article Online Lab on a Chip bated with immuno-magnetic beads and then loaded into a measurement chamber where tagged cells are drawn towards the upper surface of the measurement chamber using magnetic forces. Captured cells are then fluorescently excited using filtered LEDs and imaged using a microscope objective and CCD camera. Published on 22 April 2013. Downloaded by Dublin City University on 29/07/2013 10:07:25. Impedance-based detection strategies As well as enumerating T-helper cells by measuring the electrical impedance following CD4+ cell lysis (as used in the Daktari Diagnostics device), direct cell counting using changes in impedance have been developed.72,73 Watkins et al. developed a micro-cytometer which used an AC impedance interrogation technique to count cultured CD4+ cells with good agreement to a commercial platform across a range of concentrations.74 A similar detection strategy was then integrated into a flow-through system (Fig. 9a).75 In this case, the red blood cells are lysed off-chip and the solution containing white blood cells is loaded onto the microfluidic device. As cells flow into the chamber, they are counted by an inlet sensor. A second sensor at the exit detects cells leaving the chamber and at this point the flow is reversed. The cells are then counted again as they exit the chamber (past the inlet sensor). The difference in total cell counts indicates the number of cells trapped on the inner surface of the chamber using biotinylated anti-CD4 antibodies. A hybrid microfluidic system integrating impedance detection with fluorescent imaging was produced by Wang et al.76 Here, resistive pulse sensing (using a MOSFET) was used to count the total number of cells passing through a detection channel while simultaneously detected CD4+ cells fluorescently. This permitted the total CD4+ cell count to be expressed as a percentage of total white blood cells. Holmes et al.77,78 developed a platform to measure the electrical impedance of cells passing through a microchannel. Small antibody conjugated beads bind to CD4+ cells and modify their electrical properties such that they can be discriminated and enumerated from the leukocyte population. Along with identifying T-helper cells expressing CD4+, the platform could also successfully distinguish monocytes from lymphocytes allowing improved T-helper cell counting. Mishra et al.79 developed a system where a biosensor was composed of an electrode with dimensions of 100 mm 6 100 mm, and coated with anti-CD4+ antibody. The electrodes were electrically connected to the counter-electrode through the buffer solution. Cell binding on the electrodes generated a change in electrical impedance at low frequencies. A similar concept was developed by Jiang et al.80 In this case a biosensor composed of 200 densely packed electrodes was deposited on a microfluidic chip. Individual electrodes called pixels were sized and manufactured such that only one cell could bind on each electrode. The pixels were coated with anti-CD4+ antibody. Coated electrodes were connected to counter electrodes through a phosphate buffered saline (PBS) solution. Cell capture on the pixels showed a significant change in electrical impedance at low frequencies. Electrical counting of captured cells showed exact correlation with optical counting. This journal is ß The Royal Society of Chemistry 2013 Critical Review Other CD4+ enumerations strategies In some cases efforts have been made to take existing proven flow cytometry strategies and miniaturising them. Wang et al.81 developed a 2-stage sample preparation and cell counting system on a PDMS chip (Fig. 9b). The chip contains pneumatic chambers which can actuate the walls of the chip and hence disturb fluid through peristaltic action. During the sample preparation stage, a pneumatically actuated vortex micro-mixer is used to fluorescently label 400 ml of PBMC cells (isolated off-chip from whole blood) with anti-CD4, -CD3 and CD8 antibodies. Following incubation, a pneumatically actuated peristaltic micropump is used to drive sample into the micro-cytometer. Two additional micropumps are used to generate sheathing flows such that the cells flow past the optical detection system. This system consists of a 473-nm laser diode and three filter/photo-multiplier tubes which permit simultaneous measurement of three different wavelengths and so allowing identification of cells expressing combinations of CD3, CD4 and CD8 surface markers. Beck et al.82 have developed a microfluidic chip with a 26.5mm deep chamber but with a large platform area. The lower surface of each chamber is coated with a hydrogel functionalised with either CD3/CD4 or CD3/CD8 fluorescent antibodies. The chambers are filled with whole blood through capillary action. In the presence of whole blood, the hydrogels dissolve and uniformly release the antibodies into the sample. Slides are then imaged by using a fluorescent imaging system, and CD4+ and CD8+ cells are identified using an automated algorithm. The advantage of this approach is that diffusion through the depth of the chamber is in the order of minutes and the shallow chamber permits the tagged cells to be easily imaged. Realising that optics can be the most costly component of medical instruments; lens-less systems have been developed for cell detection in point-of-care microfluidic devices.83–87 Moon et al.88–90 developed a system in which a microfluidic substrate is positioned over a large CCD sensor. Sample is passed through the microfluidic device and CD4+ cells are bound to the floor of the microchamber using anti-CD4 antibodies. A white light source is used to illuminate the substrate from above and bound cells are counted by the diffracted shadow signal they cast onto the CCD sensor. Kiesel et al.91 have developed a spatially filtered fluorescent detection technique which can differentiate between CD4+ monocytes and lymphocytes. In this approach fluorescing bioparticles such as labelled cells flow past a large area detector. A spatial mask comprising a pseudo-random pattern means the particles only transmit in a pre-determined pattern. Correlating the detected signal with the pseudo-random pattern enables high discrimination of particles from background noise. An alternative approach to circumventing the use of highcost optics is the adaptation of smartphone based technology. Their low-cost optics and inherent connectivity and processing power offer great potential in the point-of-care arena. Zhu et al.93 adopted a cell-phone and used its high-resolution camera to detect fluorescently tagged white blood cells in dilute whole blood flowing in a 44 mm deep channel. They Lab Chip, 2013, 13, 2731–2748 | 2743 View Article Online Lab on a Chip Published on 22 April 2013. Downloaded by Dublin City University on 29/07/2013 10:07:25. Critical Review Fig. 9 Examples of alternative microfluidic strategies for CD4+ isolation and enumeration based on (a) electrical impedance measurements adapted from;75 (b) miniaturised flow cytometry used from81 with kind permission from Springer Science and Business Media; (c) label-free opto-fluidics adapted from.92 2744 | Lab Chip, 2013, 13, 2731–2748 This journal is ß The Royal Society of Chemistry 2013 View Article Online Published on 22 April 2013. Downloaded by Dublin City University on 29/07/2013 10:07:25. Lab on a Chip used particle tracking software to identify and count cells within the channel. In an alternative to fluorescence based detection, Wang et al.94 developed a microfluidic platform which estimates the number of CD4+ cells using chemiluminescence-based detection. The microfluidic chip contains microfabricated pillars which have been coated with anti-CD4 antibody. These pillars are shaped to expose cells to a range of shear stresses, such that CD4+ cells are selectively captured at the pillars but other cells flow through the system. The chamber is then washed with a chemiluminescent substrate and emitted light is monitored using a photodiode. Hydrodynamic techniques have also been used in microfluidic devices for cell-sorting and isolation by hydrodynamic size.95 Gohring and Fan92 developed a detection platform using an optofluidic ring resonator (OFRR) integrated into a microfluidic device (Fig. 9c). A fibre optic couples light into a capillary wall, where it is confined in a circular resonance known as whispering gallery mode (WGM). At specific resonant wavelengths a dip in intensity is detected at the end of the fibre-cable. This wavelength is dependent on the refractive index of the OFRR. PBS buffer containing concentrations of CD4+ cells in a medically significant range were pumped through the OFRR detection at 1 ml min21. Cells expressing CD4 and CD8 were immobilised on the inner surface of the OFRR. As they bind to the inner surface they change the refractive index of the OFRR and the subsequent shift in WGM was correlated to cell concentration. The measurement technique was also applied to CD4+ lysis. Sample lysis was found to allow more uniform arrival of sample to the sensing area and results were found to have stronger correlation to initial CD4+ cell concentrations. Cell analysis strategies with potential for CD4+ counting Microfluidic technology developed for other biomedical applications also offers the potential for adaptation to CD4+ cell capture and enumeration. Stybayeva et al.96 used surface plasmon resonance to detect interferon-gamma release from immobilised CD4+ cells, while Kobel et al.97 designed a microfluidic chip for the study of stem-cells. This consists of a microchannel with 2048 micro-traps. These traps are sized such that single cells can sediment into the trap from the flow passing over it. An image analysis algorithm is applied which can determine based on the location of the microchannel walls where the micro-traps are located, and then determines if a cell is present or not. Ibarlucea et al.98 developed a low-cost, disposable photonics chip which can use scattering and absorption for cell screening. Kim et al.99 developed a platform for isolation of circulating tumour cells (CTCs) through size filtration. In this case they artificially increase the size of CTCs by binding them to solid microbeads conjugated with anti-EpCAM, thereby making size filtration more effective. The microfluidic lab-on-a-disc platform has also offered significant potential for isolation and identification of cells.100,101 Conclusions Amongst the technologies available or under development, our critical review identified a range of candidates which have or at This journal is ß The Royal Society of Chemistry 2013 Critical Review least bear the potential to meet the full scope of the Affordable, Sensitive, Specific, User-friendly, Rapid/Robust, Equipment-Free and Delivered (ASSURED) criteria coined by the WHO for point-of-care testing in resource-poor settings. Note that a failure in just a single criterion might be prohibitive for a successful product introduction. There is typically a compromise between simplicity of operation, costs and diagnostic performance. For example, the systems loaded ‘‘un-powered’’, i.e. autonomously, through capillary action often require fluorescent detection which brings about rather complex optical setups and/or a cold-chain for the consumable/fluorophore. A cold-chain is also frequently required for systems based on cell adhesion. Systems built around a flow-cell will need a source of pumping (either integrated or external) while those systems built on impedance based detection may avoid the need for expensive fluorescence markers and optics, at the expense of manufacturing complexity and fragility of their consumable. Another criterion not included in the ASSURED criteria, but critical to emerging technologies (due to their earlier stage in the product development pipeline) is compatibility with parallel emerging technologies and trends. For example, work towards using mobile-/smart-phones for CD4+ cell enumeration have started.93 These battery-powered systems are nowadays virtually ubiquitous and widely accepted by consumers, thus bringing tremendous computational power, data communication, high-resolution displays, and integrated optics even to resource-poor settings. Similarly the development of an open diagnostic platform could make a system more affordable by permitting a single instrument to diagnose multiple diseases including HIV through CD4+ enumeration, diseases traditionally associated with developing countries such as malaria and tuberculosis; and other emerging diseases, such as diabetes, which have been identified as future problems for developing countries.102,103 Acknowledgements This work was supported by Enterprise Ireland under grant No EI - CF 2011 1317 and the Science Foundation Ireland under grant No 10/CE/B1821. Notes and references 1 F. Barré-Sinoussi, J. C. Chermann, F. Rey, M. T. Nugeyre, S. Chamaret, J. Gruest, C. Dauguet, C. Axler-Blin, F. Vézinet-Brun and C. Rouzioux, Science, 1983, 220, 868–871, DOI: 10.1126/science.6189183. 2 R. C. Gallo, P. S. Sarin, E. Gelmann, M. Robert-Guroff, E. Richardson, V. Kalyanaraman, D. Mann, G. D. Sidhu, R. E. Stahl and S. Zolla-Pazner, Science, 1983, 220, 865–867, DOI: 10.1126/science.6601823. 3 M. G. Sarngadharan, M. Popovic, L. Bruch, J. Schupbach and R. C. Gallo, Science, 1984, 224, 506–508, DOI: 10.1126/ science.6324345. 4 UNAIDS, Global Report Facet Sheet, UNAIDS, 2010. Lab Chip, 2013, 13, 2731–2748 | 2745 View Article Online Published on 22 April 2013. Downloaded by Dublin City University on 29/07/2013 10:07:25. Critical Review 5 A. Jahn, S. Floyd, A. C. Crampin, F. Mwaungulu, H. Mvula, F. Munthali, N. McGrath, J. Mwafilaso, V. Mwinuka and B. Mangongo, Lancet, 2008, 371, 1603–1611, DOI: 10.1016/ S0140-6736(08)60693-5. 6 G. Reniers, T. Araya, G. Davey, N. Nagelkerke, Y. Berhane, R. Coutinho and E. J. Sanders, AIDS, 2009, 23, 511, DOI: 10.1097/QAD.0b013e32832403d0. 7 UNAIDS, World AIDS Day Report, UNAIDS, Geneva, 2011. 8 UNAIDS, Global report: UNAIDS report on the global AIDS epidemic 2012, UNAIDS / JC2417E, WHO, Geneva, 2012. 9 M. Bofill, G. Janossy, C. Lee, D. MacDonald-Burns, A. Phillips, C. Sabin, A. Timms, M. Johnson and P. Kernoff, Clin. Exp. Immunol., 1992, 88, 243–252, DOI: 10.1111/j.1365-2249.1992.tb03068.x. 10 WHO, Antiretroviral therapy for HIV infection in Infants and children: Towards universal access, WC 503.2, World Health Organisation, Geneva, Switzerland, 2010. 11 WHO, Laboratory guidelines for enumerating CD4 T lymphocytes in the context of HIV/Aids, WH 200, WHO, New Delhi, India, 2007. 12 WHO, Rapid advice: Antiretroviral therapy for HIV infection in adults and adolescents, WHO, Geneva, 2009. 13 K. G. Castro, J. W. Ward, L. Slutsker, J. W. Buehler, H. W. Jaffe, R. L. Berkelman and J. W. Curran, Clin. Infect. Dis., 1993, 17, 802–810, DOI: 10.1093/clinids/17.4.802. 14 S. Rosen and M. P. Fox, PLoS Med., 2011, 8, e1001056, DOI: 10.1371/journal.pmed.1001056. 15 E. Losina, I. V. Bassett, J. Giddy, S. Chetty, S. Regan, R. P. Walensky, D. Ross, C. A. Scott, L. M. Uhler and J. N. Katz, PLoS One, 2010, 5, e9538, DOI: 10.1371/journal. pone.0009538. 16 I. V. Jani, N. E. Sitoe, E. R. Alfai, P. L. Chongo, J. I. Quevedo, B. M. Rocha, J. D. Lehe and T. F. Peter, Lancet, 2011, 378, 1572–1579, DOI: 10.1016/S0140-6736(11)610520. 17 C. L. Grundy, A. Medina Lara, D. Winogron, A. P. Croucher, H. G. Batz, T. B. Hallett and S. D. Reid, Sixth International AIDS Society Conference on HIV Pathogenesis, Treatment and Prevention, 2011, Abstract MOAD0105. 18 WHO, WHO Expert Meeting on short, medium and longer term product development priorities in HIV-releated diagnostics, WHO, Geneva, Switzerland, 2012. 19 N. Sitoe, E. Luecke, N. Tembe, R. Matavele, V. Cumbane, E. Macassa, P. Vaz, H. Sheppard and I. V. Jani, J. Immunol. Methods, 2011, 372, 1–6, DOI: 10.1016/j.jim.2011.07.008. 20 C. A. MacLennan, J. J. van Oosterhout, S. A. White, M. T. Drayson, E. E. Zijlstra and M. E. Molyneux, AIDS, 2007, 21, 1643, DOI: 10.1097/QAD.0b013e32823bcb03. 21 S. Mtapuri-Zinyowera, M. Chideme, D. Mangwanya, O. Mugurungi, S. Gudukeya, K. Hatzold, A. Mangwiro, G. Bhattacharya, J. Lehe and T. Peter, JAIDS, J. Acquired Immune Defic. Syndr., 2010, 55, 1, DOI: 10.1097/ QAI.0b013e3181e93071. 22 D. K. Glencross, L. M. Coetzee, M. Faal, M. Masango, W. S. Stevens, W. F. Venter and R. Osih, J. Int. AIDS Soc., 2012, 15, 3, DOI: 10.1186/1758-2652-15-3. 23 I. V. Jani, N. E. Sitoe, P. L. Chongo, E. R. Alfai, J. I. Quevedo, O. Tobaiwa, J. D. Lehe and T. F. Peter, AIDS, 2011, 25, 807, DOI: 10.1097/QAD.0b013e328344f424. 2746 | Lab Chip, 2013, 13, 2731–2748 Lab on a Chip 24 P. A. Diaw, G. Daneau, A. A. Coly, B. P. Ndiaye, D. Wade, M. Camara, S. Mboup, L. Kestens and T. N. Dieye, JAIDS, J. Acquired Immune Defic. Syndr., 2011, 58, e103–e111, DOI: 10.1097/QAI.0b013e318235b378. 25 K. Pattanapanyasat, Asian Pac. J. Allergy Immunol., 2012, 30, 11. 26 K. Pattanapanyasat, K. Sukapirom, L. Kowawisatsut and C. Thepthai, Cytometry, Part B, 2008, 74, S98–S106, DOI: 10.1002/cyto.b.20415. 27 Partec GmbH, CyFlow1 miniPOC Product Information Document, Partec GmbH, 2011. 28 M. Murtagh, 2012 HIV/AIDS Diagnostic Technology Landscape: 2nd Edition, World Health Organisation, Geneva, 2012. 29 M. Murtagh, UNITAIDS Technical Report: HIV/AIDS Diagnostic Landscape, World Health Organisation, Geneva, 2011. 30 Alere, PimaTM Analyser User Guide, Alere, Jena, 2010. 31 WHO, WHO Prequalification of Diagnostics Programme Public Report - Product: Pima CD4 Test, PQDx 0099-032-00, WHO, Geneva, 2010. 32 S. Herbert, S. Edwards, G. Carrick, A. Copas, C. Sandford, M. Amphlett and P. Benn, Sex. Transm. Infect., 2012, 88, 413–417, DOI: 10.1136/sextrans-2012-050507. 33 Y. C. Manabe, Y. Wang, A. Elbireer, B. Auerbach and B. Castelnuovo, PLoS One, 2012, 7, e34319, DOI: 10.1371/ journal.pone.0034319. 34 WHO, Technical Brief on CD4 Technologies, WHO, Geneva, 2010. 35 C. Logan, M. Givens, J. T. Ives, M. Delaney, M. J. Lochhead, R. T. Schooley and C. A. Benson, J. Immunol. Methods, 2012, 387, 107–113, DOI: 10.1016/ j.jim.2012.10.002. 36 UNITAID, 2011 HIV/AIDS Diagnostic Technology Landscape Semi-Annual Update, WHO, UNITAID, Geneva, 2011. 37 M. Murtagh, 2012 HIV/AIDS Diagnostic Technology Landscape: Semi-Annual Update, World Health Organisation, Geneva, 2012. 38 D. Klein, Small But Mighty, Continuum Inc., 2010. 39 X. Cheng, D. Irimia, M. Dixon, J. C. Ziperstein, U. Demirci, L. Zamir, R. G. Tompkins, M. Toner and W. R. Rodriguez, JAIDS, J. Acquired Immune Defic. Syndr., 2007, 45, 257–261, DOI: 10.1097/QAI.0b013e3180500303. 40 K. A. Davis, B. Abrams, S. B. Iyer, R. A. Hoffman and J. E. Bishop, Cytometry, 1999, 33, 197–205, DOI: 10.1002/ (SICI)1097-0320(19981001)33:2,197::AID-CYTO14.3.0.CO;2-P. 41 D. Collins, H. Gibson, J. Race and H. Salt, Ann. Rheum. Dis., 1939, 1, 333–358, DOI: 10.1136/ard.1.4.333. 42 D. S. Boyle, K. R. Hawkins, M. S. Steele, M. Singhal and X. Cheng, Trends Biotechnol., 2012, 30, 45–54, DOI: 10.1016/j.tibtech.2011.06.015. 43 R. Zachariah, S. Reid, P. Chaillet, M. Massaquoi, E. Schouten and A. Harries, Trop. Med. Int. Health, 2011, 16, 37–41, DOI: 10.1111/j.1365-3156.2010.02669.x. 44 H. Kettler, K. White and S. Hawkes, Mapping the landscape of diagnostics for sexually transmitted infections, TDR/STI/ IDE/04.1, WHO, UNICEF, UNDP, World Bank, Geneva, 2004. 45 D. Mabey, R. W. Peeling, A. Ustianowski and M. D. Perkins, Nat. Rev. Microbiol., 2004, 2, 231–240, DOI: 10.1038/nrmicro841. This journal is ß The Royal Society of Chemistry 2013 View Article Online Published on 22 April 2013. Downloaded by Dublin City University on 29/07/2013 10:07:25. Lab on a Chip 46 P. Balakrishnan, S. Solomon, N. Kumarasamy and K. H. Mayer, Indian J. Med. Res., 2005, 121, 345–355. 47 D. Anderson, S. Crowe and M. Garcia, Curr. HIV/AIDS Rep., 2011, 8, 31–37, DOI: 10.1007/s11904-010-0067-z. 48 A. V. Carella, M. W. Moss, V. Provost and T. C. Quinn, Clin. Diagn. Lab. Immunol., 1995, 2, 623–625. 49 P. Balakrishnan, M. Dunne, N. Kumarasamy, S. Crowe, G. Subbulakshmi, A. K. Ganesh, A. J. Cecelia, P. Roth, K. H. Mayer, S. P. Thyagarajan and S. Solomon, JAIDS, J. Acquired Immune Defic. Syndr., 2004, 36, 1006–1010, DOI: 10.1097/00126334-200408150-00002. 50 N. Watkins, D. Irimia, M. Toner and R. Bashir, IEEE Pulse, 2011, 2, 19–27, DOI: 10.1109/MPUL.2011.942762. 51 T. Peter, A. Badrichani, E. Wu, R. Freeman, B. Ncube, F. Ariki, J. Daily, Y. Shimada and M. Murtagh, Cytometry, Part B, 2008, 74B, S123–S130, DOI: 10.1002/cyto.b.20416. 52 W. G. Lee, Y. Kim, B. G. Chung, U. Demirci and A. Khademhosseini, Adv. Drug Delivery Rev., 2010, 62, 449–457, DOI: 10.1016/j.addr.2009.11.016. 53 W. R. Rodriguez, N. Christodoulides, P. N. Floriano, S. Graham, S. Mohanty, M. Dixon, M. Hsiang, T. Peter, S. Zavahir, I. Thior, D. Romanovicz, B. Bernard, A. P. Goodey, B. D. Walker and J. T. McDevitt, PLoS Med., 2005, 2, e182, DOI: 10.1371/journal.pmed.0020182. 54 J. V. Jokerst, P. N. Floriano, N. Christodoulides, G. W. Simmons and J. T. McDevitt, Lab Chip, 2008, 8, 2079–2090, DOI: 10.1039/b817116e. 55 M. A. Alyassin, S. Moon, H. O. Keles, F. Manzur, R. L. Lin, E. Haeggstrom, D. R. Kuritzkes and U. Demirci, Lab Chip, 2009, 9, 3364–3369, DOI: 10.1039/b911882a. 56 M. Hosokawa, M. Asami, S. Nakamura, T. Yoshino, N. Tsujimura, M. Takahashi, S. Nakasono, T. Tanaka and T. Matsunaga, Biotechnol. Bioeng., 2012, 109, 2017–2024, DOI: 10.1002/bit.24471. 57 T. F. Didar and M. Tabrizian, Lab Chip, 2010, 10, 3043–3053, DOI: 10.1039/c0lc00130a. 58 Y. Liu, Z. Matharu, M. Howland, A. Revzin and A. Simonian, Anal. Bioanal. Chem., 2012, 404, 1181–1196, DOI: 10.1007/s00216-012-6149-6. 59 X. Cheng, D. Irimia, M. Dixon, K. Sekine, U. Demirci, L. Zamir, R. G. Tompkins, W. Rodriguez and M. Toner, Lab Chip, 2007, 7, 170–178, DOI: 10.1039/b612966h. 60 X. Cheng, A. Gupta, C. Chen, R. G. Tompkins, W. Rodriguez and M. Toner, Lab Chip, 2009, 9, 1357–1364, DOI: 10.1039/b818813k. 61 S. Thorslund, R. Larsson, F. Nikolajeff, J. Bergquist and J. Sanchez, Sens. Actuators, B, 2007, 123, 847–855, DOI: 10.1016/j.snb.2006.10.034. 62 S. Thorslund, R. Larsson, J. Bergquist, F. Nikolajeff and J. Sanchez, Biomed. Microdevices, 2008, 10, 851–857, DOI: 10.1007/s10544-008-9199-y. 63 H. Zhu, M. Macal, C. N. Jones, M. D. George, S. Dandekar and A. Revzin, Anal. Chim. Acta, 2008, 608, 186–196, DOI: 10.1016/j.aca.2007.12.021. 64 D. Kim, J. Seol, Y. Wu, S. Ji, G. Kim, J. Hyung, S. Lee, H. Lim, R. Fan and S. Lee, Nanoscale, 2012, 4, 2500–2507, DOI: 10.1039/c2nr11338d. 65 U. A. Gurkan, T. Anand, H. Tas, D. Elkan, A. Akay, H. O. Keles and U. Demirci, Lab Chip, 2011, 11, 3979–3989, DOI: 10.1039/c1lc20487d. This journal is ß The Royal Society of Chemistry 2013 Critical Review 66 V. I. Furdui and D. J. Harrison, Lab Chip, 2004, 4, 614–618, DOI: 10.1039/b409366f. 67 D. Gao, H. Li, G. Guo and J. Lin, Talanta, 2010, 82, 528–533, DOI: 10.1016/j.talanta.2010.05.019. 68 X. Li, A. G. J. Tibbe, E. Droog, L. W. M. M. Terstappen and J. Greve, Clin. Vaccine Immunol., 2007, 14, 412–419, DOI: 10.1128/CVI.00372-06. 69 X. Li, A. Ymeti, B. Lunter, C. Breukers, A. G. J. Tibbe, L. W. M. M. Terstappen and J. Greve, Cytometry, Part B, 2007, 72B, 397–407, DOI: 10.1002/cyto.b.20165. 70 X. Li, C. Breukers, A. Ymeti, B. Lunter, L. W. M. M. Terstappen and J. Greve, Cytometry, Part B, 2009, 76B, 118–126, DOI: 10.1002/cyto.b.20445. 71 A. Ymeti, X. Li, B. Lunter, C. Breukers, A. G. J. Tibbe, L. W. M. M. Terstappen and J. Greve, Cytometry, Part A, 2007, 71A, 132–142, DOI: 10.1002/cyto.a.20375. 72 K. C. Cheung, M. Di Berardino, G. Schade-Kampmann, M. Hebeisen, A. Pierzchalski, J. Bocsi, A. Mittag and A. Tárnok, Cytometry, Part A, 2010, 77A, 648–666, DOI: 10.1002/cyto.a.20910. 73 T. Sun and H. Morgan, Microfluid. Nanofluid., 2010, 8, 423–443, DOI: 10.1007/s10404-010-0580-9. 74 N. Watkins, B. M. Venkatesan, M. Toner, W. Rodriguez and R. Bashir, Lab Chip, 2009, 9, 3177–3184, DOI: 10.1039/ b912214a. 75 N. N. Watkins, S. Sridhar, X. Cheng, G. D. Chen, M. Toner, W. Rodriguez and R. Bashir, Lab Chip, 2011, 11, 1437–1447, DOI: 10.1039/c0lc00556h. 76 Y. Wang, Y. Kang, D. Xu, C. H. Chon, L. Barnett, S. A. Kalams, D. Li and D. Li, Lab Chip, 2008, 8, 309–315, DOI: 10.1039/b713932b. 77 D. Holmes, D. Pettigrew, C. H. Reccius, J. D. Gwyer, C. van Berkel, J. Holloway, D. E. Davies and H. Morgan, Lab Chip, 2009, 9, 2881–2889, DOI: 10.1039/b910053a. 78 D. Holmes and H. Morgan, Anal. Chem., 2010, 82, 1455–1461, DOI: 10.1021/ac902568p. 79 N. N. Mishra, S. Retterer, T. J. Zieziulewicz, M. Isaacson, D. Szarowski, D. E. Mousseau, D. A. Lawrence and J. N. Turner, Biosens. Bioelectron., 2005, 21, 696–704, DOI: 10.1016/j.bios.2005.01.011. 80 X. Jiang and M. G. Spencer, Biosens. Bioelectron., 2010, 25, 1622–1628, DOI: 10.1016/j.bios.2009.11.024. 81 J. Wang, C. Wang, C. Lin, H. Lei and G. Lee, Microfluid. Nanofluid., 2011, 10, 531–541, DOI: 10.1007/s10404-0100687-z. 82 M. Beck, S. Brockhuis, N. van der Velde, C. Breukers, J. Greve and L. W. M. M. Terstappen, Lab Chip, 2012, 12, 167–173, DOI: 10.1039/c1lc20565j. 83 U. A. Gurkan, S. Moon, H. Geckil, F. Xu, S. Wang, T. J. Lu and U. Demirci, Biotechnol. J., 2011, 6, 138–149, DOI: 10.1002/biot.201000427. 84 T. Su, S. Seo, A. Erlinger and A. Ozcan, Biotechnol. Bioeng., 2009, 102, 856–868, DOI: 10.1002/bit.22116. 85 E. P. Dupont, E. Labonne, Y. Maruyama, C. Vandevyver, U. Lehmann, M. A. M. Gijs and E. Charbon, Sens. Actuators, B, 2012, 174, 609–615, DOI: 10.1016/ j.snb.2012.06.049. 86 E. D. Pratt, C. Huang, B. G. Hawkins, J. P. Gleghorn and B. J. Kirby, Chem. Eng. Sci., 2011, 66, 1508–1522, DOI: 10.1016/j.ces.2010.09.012. Lab Chip, 2013, 13, 2731–2748 | 2747 View Article Online Published on 22 April 2013. Downloaded by Dublin City University on 29/07/2013 10:07:25. Critical Review 87 A. Ozcan and U. Demirci, Lab Chip, 2008, 8, 98–106, DOI: 10.1039/b713695a. 88 S. Moon, U. A. Gurkan, J. Blander, W. W. Fawzi, S. Aboud, F. Mugusi, D. R. Kuritzkes and U. Demirci, PLoS One, 2011, 6, e21409, DOI: 10.1371/journal.pone.0021409. 89 S. J. Moon, R. Lin and U. Demirci, J. Visualized Exp., 2007, e315, DOI: 10.3791/315. 90 S. Moon, H. O. Keles, A. Ozcan, A. Khademhosseini, E. Hæggstrom, D. Kuritzkes and U. Demirci, Biosens. Bioelectron., 2009, 24, 3208–3214, DOI: 10.1016/ j.bios.2009.03.037. 91 P. Kiesel, M. Beck and N. Johnson, Cytometry, Part A, 2011, 79, 317–324, DOI: 10.1002/cyto.a.21042. 92 J. T. Gohring and X. Fan, Sensors, 2010, 10, 5798–5808, DOI: 10.3390/s100605798. 93 H. Zhu, S. Mavandadi, A. F. Coskun, O. Yaglidere and A. Ozcan, Anal. Chem., 2011, 83, 6641–6647, DOI: 10.1021/ ac201587a. 94 Z. Wang, S. Y. Chin, C. D. Chin, J. Sarik, M. Harper, J. Justman and S. K. Sia, Anal. Chem., 2010, 82, 36–40, DOI: 10.1021/ac902144w. 95 D. W. Inglis, J. A. Davis, T. J. Zieziulewicz, D. A. Lawrence, R. H. Austin and J. C. Sturm, J. Immunol. Methods, 2008, 329, 151–156, DOI: 10.1016/j.jim.2007.10.004. 2748 | Lab Chip, 2013, 13, 2731–2748 Lab on a Chip 96 G. Stybayeva, M. Kairova, E. Ramanculov, A. L. Simonian and A. Revzin, Colloids Surf., B, 2010, 80, 251–255, DOI: 10.1016/j.colsurfb.2010.06.015. 97 S. A. Kobel, O. Burri, A. Griffa, M. Girotra, A. Seitz and M. P. Lutolf, Lab Chip, 2012, 12, 2843–2849, DOI: 10.1039/ c2lc40317j. 98 B. Ibarlucea, E. Fernandez-Rosas, J. Vila-Planas, S. Demming, C. Nogues, J. A. Plaza, S. Büttgenbach and A. Llobera, Anal. Chem., 2010, 82, 4246–4251, DOI: 10.1021/ac100590z. 99 M. S. Kim, T. S. Sim, Y. J. Kim, S. S. Kim, H. Jeong, J. Park, H. Moon, S. I. Kim, O. Gurel, S. S. Lee, J. Lee and J. C. Park, Lab Chip, 2012, 12, 2874–2880, DOI: 10.1039/ c2lc40065k. 100 R. Burger, D. Kirby, M. Glynn, C. Nwankire, M. O’Sullivan, J. Siegrist, D. Kinahan, G. Aguirre, G. Kijanka, R. A. Gorkin III and J. Ducrée, Curr. Opin. Chem. Biol., 2012, 16, 409–414, DOI: 10.1016/j.cbpa.2012.06.002. 101 R. Burger and J. Ducrée, Expert Rev. Mol. Diagn., 2012, 12, 407–421, DOI: 10.1586/erm.12.28. 102 P. Hossain, B. Kawar and M. El Nahas, N. Engl. J. Med., 2007, 356, 213–215, DOI: 10.1056/NEJMp068177. 103 A. M. Prentice, Int. J. Epidemiol., 2006, 35, 93–99, DOI: 10.1093/ije/dyi272. This journal is ß The Royal Society of Chemistry 2013