Development of microfluidic devices for the separation of blood plasma from capillary samples

Abstract

Point-of-care devices have the potential to revolutionise healthcare as we know it. Sensing platforms capable of performing clinical tests rapidly and at the site of a patient can facilitate early diagnoses and enable continuous patient care in chronic conditions. This has been clearly demonstrated by glucose monitoring devices, which are now the standard of care for patients with diabetes. However, with the exception of glucose sensors and lateral flow devices such as those used in pregnancy or Covid-19 testing, point-of-care devices are not commonly used in modern medicine. One of the main reasons for this is the difficulty in separating blood cells from plasma, the liquid part of blood, without damaging red blood cells and, consequently, releasing haemoglobin in the plasma. The absence of haemoglobin in plasma samples is a strict requirement for many clinical tests and diagnostic procedures. Microfluidic technologies offer an array of tools for understanding and controlling very small volumes of fluids within microchannels. They are a key part of most lab-on-a-chip and biosensing systems used in point-of-care testing and can be successfully used to gently separate blood cells from plasma, thus preparing a sample for analysis.

The microfluidic devices for sample preparation proposed in the literature often require too many complex manual steps to assemble and use, are poorly analysed, can be ineffective under normal clinical circumstances or are too expensive to produce and commercialise. This research seeks to ll the gap in the literature for thoroughly characterised, low-cost, passive microfiltration devices operated without power or specialist equipment for the separation of good quality blood plasma from red blood cells in undiluted capillary samples ≤ 100 μL. Two main rapid manufacturing techniques were used and compared, both of which allow the plasma separation devices to be potentially incorporated into, or expanded to become full point-of-care devices: laser cutting with subsequent bonding of a thermoplastic material and 3D printing, the latter using a plant-based biodegradable material and an affordable printer. Several side studies were carried out to assess the suitability of these methods for the manufacturing of microfluidic device prototypes. All iterations of the devices developed were tested extensively with a wide range of whole blood samples and their performance was evaluated by analysing the percentage of available plasma extracted, the time necessary for extraction and the device failure rate. To determine the quality of the plasma collected and its suitability for clinical testing, its haemoglobin concentrations were measured and compared with the concentrations found in control plasma samples prepared using gold standard techniques.

The laser-cut devices were operated without power or specialist equipment, requiring only a commercial metered device used for the collection of blood from a finger prick to actuate the blood flow in a dead-end fi ltration setup. The best iterations of these devices could extract on average 54.85% of the available plasma volume from 100 μL of undiluted whole blood in three minutes, with only 7.89%. failing during the testing process. The high concentrations of haemoglobin found in the plasma, however, made the devices unsuitable for a wide range of diagnostic tests.

The 3D printed devices only required a pipette for their operation, with the best iteration recovering on average 56.88% of the total available plasma from 50 μL whole blood samples in 87 seconds. The quality of the extracted plasma was excellent, with a negligible haemoglobin concentration difference with control samples. The plasma collected using the 3D printed devices was 99.9% pure and was tested for Bovine Respiratory Syncytial Virus assay, with the results showing no discernible difference with control samples. The 3D printed devices are easy to manufacture and assemble, with some iterations being reusable after disinfection owing to their simple snap- fit mechanism. No waiting time is necessary for their operation, as only one step is required for a successful extraction. Their designs are parametric and therefore easily scalable and adjustable to accommodate for different microfilters and pipette tips. The possibility of creating fully 3D printed actuating components to integrate in 3D printed microfluidic devices was also briefly explored.

In summary, this project demonstrated how simple rapid manufacturing techniques can be used to develop low cost, yet functional, microfluidic sample preparation modules that require minimal operational steps while also being easy to manufacture and assemble. The devices can be used by the research community to help with the development of biosensing platforms, as they provide a simple and well tested template that can be easily incorporated in lab-on-a-chip setup.