Citation: Bouillet A, Girois L, Weidenhaupt M, “Enhancing Protein Preservation During Biological Drug Preparation & Delivery”. ONdrugDelivery, Issue 113 (Oct 2020), pp 46–48.

Adrien Bouillet, Loïc Girois and Marianne Weidenhaupt, discuss the results of work carried out in joint efforts by EVEON and LMGP in the field of protein adsorption and aggregation, and how the results of this work have led to improvements in the design of EVEON’s proprietary micropump.

Therapeutic proteins are injectable biological drugs with a high specificity, which are used for the treatment of many diseases, including diabetes, autoimmune diseases and cancer. They are regularly marketed as lyophilised powders that need to be reconstituted, either manually or automatically, prior to injection.

“It is well established that exposure to certain material and air interfaces, to which proteins adsorb, can have a strong impact on their stability.”

It is well established that exposure to certain material and air interfaces, to which proteins adsorb, can have a strong impact on their stability.1 While adsorbing at interfaces, proteins change their conformation, which can lead to the formation of protein aggregates. This can entail a loss of function or lead to the development of immunogenic responses and, consequently, adverse reactions in patients. Moreover, these aggregates can also obstruct the fluid flow in reconstitution and injection devices, having a negative impact on their performance. Improving the stability of these drugs is generally achieved by formulation excipients, such as surfactants, but can also be maximised by a precise design of device components and fluidic protocols.

In order to enhance protein preservation during drug preparation and delivery, EVEON and the Laboratoire des Matériaux et du Génie Physique (LMGP) have worked together through the LabCom programme to improve EVEON’s technology platform. Three different versions of EVEON’s proprietary micropump were studied for their mechanical performance, analysing the stability of an unformulated drug solution during pump cycles at different speeds (20, 50 and 80 rotations per minute).

The drug chosen was an unformulated human insulin solution at 0.7 mg/mL in a 25 mM tris buffer of pH 7.4 with 125 mM NaCl. Unformulated insulin is well known for its high tendency to adsorb and aggregate at interfaces,2 and can therefore be considered a suitable test case for a worst-case scenario with respect to drug aggregation. The unformulated protein solution was transferred through the pumps over 500 suction/discharge cycles. The operation specifications define 300 pumping cycles as a target limit. Typically, the torque values increase with pump cycles at all speeds.

Figure 1 shows the correlation between torque increase and speed for EVEON’s original micropump and comparative results for the three pumps are shown at 80 rotations per minute in Figure 2. The micropump evolution (siliconised) showed the lowest and most stable torque values.

Figure 1: Torque as a function of number of cycles at three different speeds for EVEON’s original micropump.

Figure 2: Torque as a function of number of cycles at 80 rotations per minute for three different versions of EVEON’s proprietary micropump: original, evolution (silicone-free) and evolution (siliconised).

Insulin stability was monitored by Thioflavin T fluorescence (480 nm), a conformation-sensitive dye indicative of amyloid aggregates. The appearance of Thioflavin T-positive insulin aggregates in unformulated solutions was recorded for all pumps. The extent of fluorescence depended on insulin concentration, pumping speed and the number of cycles through the pumps. A thorough comparative analysis was completed on the micropump evolution (silicone free) and the micropump evolution (siliconised), the two pumps that showed the lowest torque values. At 80 rotations per minute and 500 cycles, aggregating was detected in one out of nine experimental runs for the micropump evolution (siliconised), whereas it was detected in four out of nine for the micropump evolution silicone free (Figures 3 and 4).

Figure 3: Insulin aggregation after 500 cycles at 80 rotations per minute for the two evolution versions of the pump. The number of experiments with Thioflavin T-positive insulin aggregation is shown with blue bars, and the total number of experiments (nine) is shown with grey bars.

Figure 4: An experiment is considered ThioflavinT-positive when the fluorescence increment after 50 pump cycles is greater than five, and the fluorescence signal at 500 cycles is greater than seven times the baseline fluorescence.

Figure 5: EVEON’s patented micropump.

These results show that reduced torque values correlate positively with a lower aggregation potential when tested with unformulated insulin solutions in the pumps. All EVEON’s micropumps showed no Thioflavin T-positive insulin aggregates when tested with formulated insulin solutions. The results led to a better understanding of the original version of the micropump, which allowed for a drastic reduction of the aggregation potential with the design of the new micropump evolution and its siliconised version.

In conclusion, the results confirmed that the new micropump evolution, even in its silicone-free version, may be considered a preferable solution when looking to mitigate the risk of protein aggregates in biopharmaceuticals during drug preparation and delivery.

EVEON’s proprietary micropump (Figure 5) is a modular platform design that may offer a reliable solution to pharmaceutical and biotech companies aiming to develop an automated drug preparation device for at-home patient care.

The authors would like to thank Lydia Esteban-Enjuto, Mahutin Akle, Eline Lopez-Soler, Abdallah Alhalabi, Benjamin Auvray, Antoine Maze and Franz Bruckert for their valuable contributions.


  1. Pinholt C et al, “The importance of interfaces in protein drug delivery-why is protein adsorption of interest in pharmaceutical formulations?”. Expert Opin Drug Deliv, 2011, Vol 8(7), pp 949–964.
  2. Sluzky V et al, “Kinetics of insulin aggregation in aqueous solutions upon agitation in the presence of hydrophobic surfaces”. Proc Natl Acad Sci USA, 1991, Vol 88(21), pp 9377–9381.

Comment on this article