The technical field generally relates to devices and methods used microscale chemical reactions. More specifically, the technical field generally relates to miniaturized or microfluidic platforms for performing chemical reactions and in particular radiochemical reactions.
Despite the increasing importance of positron emission tomography (PET) imaging in biological and clinical research, access to myriad new radioactive tracers is limited. This is in part due to the short half-life, which requires daily production close to the site of use, and also due to the high complexity of tracer production, which requires costly equipment operated behind radiation shielding and other safety infrastructure. Miniaturization of the equipment via digital microfluidic radiosynthesizer technology reduces the cost of the automated synthesizer, radiation shielding and reagents. This could enable a move away from the centralized production model, where cost is reduced by making large batches of a small number of tracers such that the costs of each can be spread over many end users. By instead making it affordable to make smaller batches of tracers, microscale tracer production could be practiced directly by the end user, increasing access to diverse tracers.
Microfluidics have been widely used in radiochemical filed. Many investigators have been working on microfluidic radiosynthesizers based on flow-through (or “continuous flow”) microfluidics, where reactions occur by flowing reagent streams through mixers and capillary tubes or other structures. Flow-through synthesizers have been commercialized (e.g., Advion Biosciences). The flow-through based radiosynthesizer can significantly reduce the precursor concentration needed for reactions compared to macroscale radiosynthesizer, and can sometimes improve reaction speed or yield. However, the size of the synthesizer is still very large (comparable in size to traditional macroscale systems), thus not providing any advantages in terms of instrument cost, shielding cost, or lab real estate used. Furthermore, since evaporation steps cannot be performed in microchannels or capillaries of flow-through systems, these reactors must be coupled with bulky conventional apparatus to perform these steps.
The electro-wetting on dielectric (EWOD) platform addresses some of these shortcomings. All steps of radiochemical reactions can be performed on the same EWOD chip, including evaporation, mixing and reactions, and only tiny volumes of reagents are needed. The electrowetting on-dielectric (EWOD) microfluidic platform has been used to efficiently synthesize several tracers in a compact form factor. Furthermore, the tracers are produced with minimal reagent consumption and have very high “specific activity.” See Chen et al., Radiolabelling Diverse Positron Emission Tomography (PET) Tracers Using a Single Digital Microfluidic Reactor Chip,” Lab on a Chip, Vol. 14, pp. 902-910 (2014). Nonetheless, the wide spread adoption of the EWOD platform is limited by the high cost for the EWOD chip and its complicated operation. EWOD chips are fabricated from several layers, including an electrode layer (e.g., indium-tin-oxide or ITO, gold, or other metals) for droplet actuation electrodes and for connections to the external control system, a dielectric layer to prevent current flow through the reaction droplet, and a hydrophobic layer. The large number of processing steps makes the EWOD chips expensive and the relatively large surface area (e.g., ˜25 mm square) makes it challenging to produce the pinhole-free dielectric layers that are essential to avoid dielectric breakdown and electrolysis of droplets on the chip.
Other investigators have looked at passive droplet manipulation that do not require electrodes or other means of actuation. The majority of reported methods for passive droplet manipulation are driven by gradients in geometry or surface tension. For example, Yeh et al. described a method to generate a gradient in the density of hydrophobic decyltrichlorosilane (DTS) molecules on the substrate. This chemical gradient on the surface leads to an imbalance in surface tension and resulting in a net force on the droplet toward the more hydrophilic side (i.e., lower DTS density). See Yeh S I et al., Droplets coalescence and mixing with identical and distinct surface tension on a wettability gradient surface, Microfluidics and Nanofluidics, 14:785-795 (2013). Ng et al. reported a method to manipulate droplets using Marangoni force. An ethanol droplet is used to create a surface tension gradient in an adjacent water droplet via evaporation of ethanol that then mixes into this second droplet. The ethanol concentration is highest on the surface of the water droplet closest to the ethanol droplet and lowest on the opposite side of the water droplet. The difference in surface tension drives the water droplet move to the opposite direction where the ethanol droplet is positioned on the hydrophilic track. See Vi-Vie Ng et al., Marangoni-induced actuation of miscible liquid droplets on an incline, In International Journal of Multiphase Flow, V. 82, pp. 27-34 (2016). Hong et al. studied the droplet transportation between two non-parallel electrode plates via AC electrowetting. The surface tension difference caused by non-parallel plates works together with interfacial oscillation driven by AC electrowetting and pulls the droplet to move toward narrow side. See J. Hong et al., Drop transport between two non-parallel plates via AC electrowetting-driven oscillation, Sens. Actuators B Chem., 188, pp. 637-643 (2013). Xing et al. reported that droplets can also be passively manipulated through capillary micropumping. A droplet of smaller volume is pumped along a superhydrophilic pathway on a hydrophobic surface to the side where a larger liquid droplet is located by a pressure difference across the droplet interface. See S. Xing et al., Droplet-driven transports on superhydrophobic-patterned surface microfluidics, Lab. Chip, 11, pp. 3642-3648 (2011).
While these techniques provide a wide range of possible transport mechanisms, not all would be suitable for performing multi-step chemical reactions. The presence of solvents or surface molecules used to provide chemical gradients could potentially interfere with, or be affected by, the intended chemical reactions on the chip. Although fabrication processes are simple in some cases, approaches such as gradient of DTS density and gradient of height do not appear to lend themselves to the creation of sophisticated channel networks.
In one embodiment, a passive microfluidic chip is provided that includes a patterned surface based on “patterned wettability,” whereby droplets are passively manipulated, thus eliminating the need for two of the three functional layers of the EWOD chip (i.e., the electrode and dielectric layers). The passive microfluidic chip operates on passive transport, whereby droplets travel along tapered hydrophilic channels or pathways (e.g., tracks) that are formed on a hydrophobic substrate. When a droplet is placed on a tapered hydrophilic channel or pathway, the contact line at the wide end of the taper is longer and has a higher total force than the contact line at the rear of the droplet. This force imbalance naturally drives droplets toward the wide end of the hydrophilic channel or pathway. In the microfluidic chip described herein, a plurality of such tapered tracks are located on the microfluidic chip and lead to a common or central region on the microfluidic chip that is used as a reaction site. For example, the microfluidic chip includes a central reaction site with a plurality of tapered tracks that radiate outward from the central reaction site. With such ability, combined with an optional temperature control mechanism, it is possible to perform all of the unit operations needed for multistep radiochemical reactions, including the movement of reagents from fixed dispensers to a central reaction site, performing evaporations, and performing reactions.
In one embodiment, a passive microfluidic device includes a hydrophobic substrate having patterned thereon a hydrophilic reaction site and a plurality of hydrophilic channels or pathways extending outward from the hydrophilic reaction site. One or more of the hydrophilic channels or pathways may terminate at respective loading sites on the substrate. The hydrophilic channels or pathways are tapered with an increasing width in an inward direction from the loading sites to the hydrophilic reaction site.
In another embodiment, a passive microfluidic device includes a hydrophobic substrate having patterned thereon a hydrophilic reaction site and one or more hydrophilic channels or pathways extending outward from the hydrophilic reaction site on the substrate, wherein the one or more hydrophilic channels or pathways are tapered along a length thereof and have a taper angle of at least 4°. The one or more hydrophilic channels or pathways may be straight or curved. In some embodiments, the one or more hydrophilic channels or pathways terminate at loading sites where droplets may be loaded. In other embodiments, the one or more hydrophilic channels or pathways terminate at a junction or the like whereby multiple droplets can be combined or mixed into a single droplet that can then traverse the hydrophilic channel or pathway to the hydrophilic reaction site.
In another embodiment, a microfluidic synthesis platform includes a microfluidic chip holder comprising a computer controlled heating element and cooling element. A microfluidic chip is mountable in the microfluidic chip holder, the microfluidic chip being formed from a hydrophobic substrate having patterned thereon a hydrophilic reaction site and a plurality of hydrophilic channels or pathways extending outward from the hydrophilic reaction site and terminating at respective loading sites on the substrate, wherein the hydrophilic channels or pathways are tapered with an increasing width in an inward direction from the loading sites to the hydrophilic reaction site. The microfluidic synthesis platform includes a fixture holding a plurality of non-contact reagent dispensing devices disposed above the microfluidic chip at locations corresponding to the loading sites of the plurality of hydrophilic channels or pathways, the fixture further holding a moveable collection tube disposed above the hydrophilic reaction site of the microfluidic chip.
In yet another embodiment, a method of using the microfluidic synthesis platform described herein includes loading one or more liquid droplets containing reactants onto one or more of the plurality of hydrophilic channels or pathways of the microfluidic chip via one or more of the non-contact reagent dispensing devices, whereby the liquid droplets spontaneously move to the hydrophilic reaction site; and reacting the reactants at the hydrophilic reaction site.
In still another embodiment, a method of synthesizing [18F]fallypride using the microfluidic synthesis platform includes loading one or more droplets containing [18F]fluoride solution onto of the hydrophilic channels or pathways, wherein the droplet containing [18F]fluoride solution automatically transports to the hydrophilic reaction site; heating the microfluidic chip with the heating element to evaporate the droplet(s) containing [18F]fluoride solution; loading one or more droplets containing fallypride precursor solution at another loading site, wherein the droplets containing fallypride precursor solution automatically transports to the hydrophilic reaction site; heating the microfluidic chip with the heating element to form [18F]fallypride; loading a plurality of droplets of a diluting solution at another loading site, wherein the droplets containing the dilution solution automatically transports to the hydrophilic reaction site containing the [18F]fallypride; and removing the droplet from the microfluidic chip (e.g., transferring a droplet containing the [18F]fallypride to a vial, container via the collection tube, or other downstream process).
In yet another embodiment, a method of synthesizing [18F]FDG using the microfluidic synthesis platform includes loading one or more droplets containing [18F]fluoride/K222/K2CO3 solution onto of the hydrophilic channels or pathways, wherein the one or more droplets containing [18F]fluoride/K222/K2CO3 solution automatically transports to the hydrophilic reaction site; heating the microfluidic chip with the heating element to evaporate the one or more droplets containing [18F]fluoride/K222/K2CO3 solution; loading a plurality of droplets (or one or more droplets) containing mannose triflate solution at another loading site, wherein the droplets containing mannose triflate solution automatically transports to the hydrophilic reaction site; heating the microfluidic chip with the heating element; loading one or more droplets of sodium hydroxide solution at another loading site, wherein the one or more droplets containing sodium hydroxide solution automatically transports to the hydrophilic reaction site whereby a hydrolysis reaction takes place to generate [18F]FDG; loading a plurality of droplets of a diluting solution at another loading site, wherein the droplets containing the dilution solution automatically transports to the hydrophilic reaction site; and removing the droplet from the microfluidic chip (e.g., transferring a droplet containing the [18F]FDG to a vial, container, or other downstream process via the collection tube).
Each of the hydrophilic channels or pathways 18 are tapered with an increasing width as one proceeds in an inward direction to the hydrophilic reaction or collection site 16. The hydrophilicity of the exposed silicon features (e.g., reaction or collection site 16 and channels or pathways 18) may be enhanced by exposing the same to a Piranha cleaning solution as explained herein during the manufacturing process. In some embodiments, the hydrophilic reaction or collection site 16 and the hydrophilic channels or pathways 18 may be rendered or made superhydrophilic while the hydrophobic layer 14 may be superhydrophobic.
In one embodiment, the central hydrophilic reaction or collection site 16 has a diameter of around 2.0-3.0 mm. As seen in
With reference to
In some embodiments, the surface of the passive microfluidic chip 10 may be open or exposed to the external environment (i.e., there is no top) as is illustrated in the cross-sectional view of
To fabricate the chip as seen in
For automated droplet reactions, a custom-built temperature controlled synthesis platform 60 was used to control the dispensing of reagents and the collection of the final product.
The chip holder 62 was made as a plastic frame above the Peltier cooling device 66 (and bolted to the heatsink 68 via bolt 63) helped keep the heater 64 in place while also providing two flat vertical edges E for rapidly positioning one corner of the microfluidic chip 10. The signal from a K-type thermocouple embedded in the heater 64 was amplified through a K-type thermocouple amplifier 65 (
With reference to
Reagent stock solutions can be pipetted directly into the vial 78. For precursor solution and [18F]fluoride solution, the smaller volume (30-50 μL) was loaded into a 250 μl vial insert (5181-1270, Agilent Technologies, Santa Clarita, Calif., USA) installed into the V-vial. The outlet of each dispenser 76 was fitted with a nozzle (ID 0.005″, INZA4650935K, Lee Company), which is recommended for generation of droplets with volume in the range of 100s of nL to several μL. With reference to
While the above describes the [18F]fluoride solution being loaded into a vial or other container that is connected to the dispenser 76 it should be appreciated that the [18F]fluoride solution may also come directly from the cyclotron, a fluoride dispenser, a fluoride concentrator, or any other source. In such configurations a vial may not be needed for fluoride loading as the [18F]fluoride solution may come from tubing, nozzle, or other dispenser.
A fixture 90 was built to hold six (6) dispensers 76 with nozzles ˜2 mm above the six (6) loading sites of the microfluidic chip 10 (only one such dispenser 76 is illustrated in
A liquid collection subsystem was implemented in the microfluidic synthesis platform 60 to transfer the final crude reaction product droplet 100 from the microfluidic chip 10 to the collection vial. As seen in
Prototype passive microfluidic chips 10 were first used to test the feasibility of multi-step radiochemical reactions. In a first example, the synthesis of [18F]fallypride, a tracer for brain imaging used in PET and related to neuropsychiatric conditions and aging, was performed on a passive microfluidic chip 10 that was loaded manually. In this first experiment, the prototype passive microfluidic chip 10 included a single hydrophilic reaction or collection site 16 and a single hydrophilic channel or pathway 18 (i.e., “one track chip”). As explained below, an additional manual experiment was carried out on a passive microfluidic chip 10 that included a plurality of hydrophilic channels or pathways 18. The synthesis scheme of [18F]fallypride is seen in
Performance was assessed via measurements of radioactivity (via dose calibrator) at the beginning and ending of each step. Fluorination efficiency of the extracted product was measured via radio thin layer chromatography (radio-TLC, Minigita Star detector, BakerFlex IB-F silica gel plates, mobile phase 60% MeCN in 25 mM ammonium formate aqueous solution with 1% TEA). Crude radiochemical yield (crude RCY) is fluorination efficiency multiplied by extracted radioactivity, divided by starting radioactivity. Distribution of radioactivity on the passive chip after each step was assessed via Cerenkov imaging.
[18F]fallypride synthesis on the prototype (“one track chip”) passive microfluidic chip exhibited 89±2% fluorination efficiency (n=3) and 60±3% crude RCY (n=3), which are comparable to the performance on an EWOD chip (crude RCY, 84±7%, n=6). Activity loss at each step was calculated for future protocol optimization, all details of performance were summarized in Table 2 below.
The distribution of radioactivity at various stages verified that majority of activity was moved to the reaction site, and minimal residual radioactivity was left in the reaction site after extraction.
[18F]fallypride was also synthesized without pipetting using the automated microfluidic synthesis platform 60 illustrated in
The performance of [18F]fallypride synthesis for both manual synthesis and automated synthesis is summarized in Table 3 below.
With manual operations, the fluorination efficiency was 74±8% (n=4), collection efficiency was 90±4% (n=4), and the crude RCY was 59±9% (n=4). Analysis of radioactivity measurements during the synthesis on passive chips showed negligible losses (˜3±1%, n=4, relative to the starting radioactivity) during drying of [18F]fluoride, but slightly higher losses of 15±2% (n=4) during fluorination and 9±4% (n=4) residual activity on chip and pipette tips after collection. Note that the negative evaporation loss is likely due to measurement error in the dose calibrator. The crude RCY was slightly lower than previously reported for the droplet-based synthesis using EWOD chips, i.e. 84±7% (n=6). The reported fluorination and collection efficiencies on EWOD were 90±9% (n=6) and 94±3% (n=6), respectively, suggesting the current platform and reaction conditions give slightly lower fluorination efficiency. The synthesis time (up to the end of the collection process) for [18F]fallypride was ˜25 min.
Automated loading and collection provided a marginal increase in the crude RCY of [18F]fallypride to 64±6% (n=4). The isolated RCY was 46±4% (n=4). Typically, in macroscale synthesis, about 5-10% of the radioactivity of the crude product can be lost during purification and formulation. Here about 28% of the activity was lost, suggesting that significant improvements can be made, perhaps in injection of the small volume of collected product into the HPLC system. The synthesis time was reduced to ˜20 minutes (˜12 minutes for drying and fluorination steps and ˜8 minutes for collection) due to elimination of manual steps. This time is slightly shorter than reported for EWOD-based synthesis (i.e., 31 minutes for [18F]fallypride). The synthesis time is also somewhat shorter than macroscale processes (˜29 minutes for [18F]fallypride). It should be pointed out that purification and formulation times were omitted for comparison purposes, but it is expected that these processes could be performed more quickly than in the reported macroscale methods due to the smaller crude reaction product volume. This smaller volume can accelerate both the purification and formulation processes. The time for collection can possibly be further condensed by optimizing the speed of the product droplet collection process.
[18F]FDG is analogue of glucose and is by far the most commonly used PET imaging tracer. In an initial experiment, the fluorination of [18F]FDG was performed using a microfluidic chip 10 having multiple tracks (e.g., like that illustrated in
[18F]FDG was also synthesized using the automated microfluidic synthesis platform 60 illustrated in
[18F]FDG was further purified with a custom miniaturized cartridge adapted from a commercially available FDG Purification cartridge for the base hydrolysis (Chromabond Set V, ABX). The commercial cartridge was designed for macroscale purification of [18F]FDG and would result in too much dilution of the purified product when making small amounts of the tracer. Instead, custom cartridges were made by repacking the resin beads (18.4 mg cation exchange resin (PS-H+), 18.2 mg anion exchange resin (PS-HCO3), 30.4 mg neutral alumina (ALOX N) and 15.6 mg reversed-phase resin (HR-P)) inside a 0.063″ ID perfluoroalkoxy alkane (PFA) tubing (ZEUS, Orangeburg, S.C., USA). The resins were sandwiched and separated with ˜1.5 mm diameter fits (FRPE1CC, OROCHEM, Naperville, Ill., USA). Before use, the cartridge was pre-conditioned with 0.5 mL EtOH and then 1 mL DI water. During purification, the ˜100 μL diluted crude product was passed through the cartridge, then an additional 300 μL DI water were used to collect the pure product.
The performance of [18F]FDG synthesis (for both full manual synthesis and automated synthesis) is shown in Table 4 below.
With manual handling, the fluorination efficiency, collection efficiency and crude RCY were 84±4% (n=4), 70±15% (n=4) and 40±8% (n=4), respectively. The intermediate [18F]FTAG was completely hydrolyzed into [18F]FDG as determined by radio-TLC with hexane and ethyl acetate (1:1, v:v) mobile phase. Evaporation loss (1±2%, n=4) was minimal, but some radioactivity losses were observed during fluorination (30±3%, n=4), and there was significant residual radioactivity on the chip and pipette tips after collection (20±9%, n=4). These losses were comparable to the losses reported on EWOD platform (48±3%, n=2). The crude RCY was marginally lower than crude RCY on EWOD platform (45±10%, n=2), which was mainly due to lower fluorination efficiency compared to that on EWOD platform (93±3%, n=2). After fluorination, the radioactivity would ideally be uniformly distributed through the reaction droplet, but in fact is confined to a smaller region corresponding to the location of the [18F]TBAF residue after the initial [18F]fluoride drying step. Further optimizations will focus on improving fluorination efficiency by adding external means to facilitate mixing and dissolution processes, and by carefully controlling volume of solution in reaction site.
As expected, after implementation of automated reagent loading and product collection for [18F]FDG synthesis, the crude RCY was enhanced to 50±8% (n=4). Higher radioactivity recovery (69±5%, n=4) compared to manual synthesis (49±12%, n=4) dominated the increase of crude RCY. Isolated RCY was 36±6% (n=4). Purification efficiency of the custom cartridge was 72±9% (n=4) and can likely be improved by optimization of the custom purification cartridges and purification process. The synthesis time (up to the end of the collection process) for [18F]FDG was ˜21 minutes, which is significantly shorter than reported for EWOD-based synthesis (˜50 min). Including purification, the synthesis time was ˜36 minutes, which only slightly longer than the fastest times reported for macroscale synthesis (˜22 min).
The starting activity of [18F]fallypride and [18F]FDG synthesis was minimized in preliminary experiments for safety reasons to ˜0.2-0.4 mCi by loading a 2 μL droplet (radioactivity concentration ˜0.1-0.2 mCi/μL). Though sufficient radioactivity was recovered for small animal imaging, it will be desirable in the future to scale this up to enable tracer production for multiple animal studies or for clinical doses. One way to scale up activity is by pre-concentrating the [18F]fluoride solution from the cyclotron using a miniaturized anion exchange cartridges. Another approach is to repeatedly load droplets of [18F]fluoride solution before drying.
Preliminary experiments were conducted to test the feasibility of the latter approach with manual loading and collecting. To ensure the same ratio of precursor to TBAHCO3 in the fluorination reaction, multiple 2 μL droplets of [18F]fluoride/[18O]H2O solution (i.e. no TBAHCO3) were first loaded, followed by a single 2 μL, droplet of TBAHCO3 solution (3.6 mM). The mixture was dried at 105° C. for 1 min. In a preliminary experiment, loading of 2, 4, or 8 μL, of [18F]fluoride/[18O]H2O solution resulted in crude RCYs was 58% (n=1), 54% (n=1), and 50% (n=1). Results are shown in Table 5 below.
This apparent reduction in crude RCY as a function of [18F]fluoride/[18O]H2O volume could be a mixing issue as it may become increasingly difficult to dissolve the increasing amount of residue in the precursor solution droplet prior to fluorination. Instead of loading a large volume and drying it once, each droplet of [18F]fluoride/[18O]H2O could be dried after loading, perhaps limiting the lateral extent of the initial residue. For the 8 μL case, the starting activity was 2.34 MBq (˜0.63 mCi).
Though the volume capacity of the chip used in the experiments was low (e.g., a few microliters), it is likely that the reagent pathways and reaction region could be scaled up or down to accommodate larger or smaller reaction volumes, respectively. Another way to scale-up is to combine with an auxiliary miniature ion-exchange cartridge, which allows large amounts of radioactive isotope to be concentrated into microliter volumes prior to loading onto the microfluidic chip 10. Thus, a concentrated droplet containing reactants can be used in conjunction with the microfluidic chip 10. This would allow production of sufficient quantities of tracers for clinical PET imaging. In addition to the production of radiotracers for PET, the chip could also be used to make agents for Single-Photon Emission Computed Tomography (SPECT) imaging, labeled peptides/proteins, or radiolabeled therapeutic molecules. It could be used to perform small scale chemical reactions in a variety of other applications as well.
The passive microfluidic device 10 has numerous advantages over EWOD-based devices and other microfluidic approaches. First, microscale radiochemical reactions largely reduce the cost of reagents. Using microliter scale reactions, <1% of the amount of reagents used for macroscale reactions are needed while maintaining similar or higher concentrations. The fabrication of passive chips is much less complicated and much less expensive than EWOD chips or other microfluidic methods for performing chemical reactions. Passive microfluidic chips 10 eliminate the need of extra electronics and pumps to manipulate the fluids on the chip. In addition, the compact size of the passive microfluidic platform requires much less shielding than a traditional macroscale radiosynthesizer. While the latter must be located in a hot cell weighing several tons, the chip can be shielded with the same thickness and potentially be light enough in weight to be used on the benchtop.
While the microfluidic chip 10 described herein has generally been described as containing a single hydrophilic reaction or collection site 16 it should be appreciated that the microfluidic chip 10 may contain multiple hydrophilic reaction or collection sites 16 and associated delivery channels or pathways 18 on a single substrate 12. For example, the multiple hydrophilic reaction or collection sites 16 may be arranged as an array such that parallel or multiple separate synthesis reactions can take place on a single microfluidic chip 10. Such an embodiment, may require a moveable reagent dispenser and/or moveable microfluidic chip 10 (e.g., using a moveable stage or the like) whereby fluids may be loaded and unloaded from the various locations formed on the microfluidic chip.
While embodiments of the present invention have been shown and described, various modifications may be made without departing from the scope of the present invention. The invention, therefore, should not be limited, except to the following claims, and their equivalents.
This Application claims priority to U.S. Provisional Patent Application No. 62/405,758 filed on Oct. 7, 2016, which is hereby incorporated by reference in its entirety. Priority is claimed pursuant to 35 U.S.C. § 119 and any other applicable statute.
This invention was made with Government support under Grant Nos. AG049918, MH097271, CA212718, and CA016042 awarded by the National Institutes of Health. The Government has certain rights in the invention.
Filing Document | Filing Date | Country | Kind |
---|---|---|---|
PCT/US2017/055585 | 10/6/2017 | WO | 00 |
Number | Date | Country | |
---|---|---|---|
62405758 | Oct 2016 | US |