Design and fabrication of ¹²⁵I seeds for brachytherapy using capillary-based microfluidic technique

The experimental facilities includes capillaries with hydrophilic (fused silica tubing, Polymicro Technologies, INNOSEP Company, Zheng-Zhou, China) or hydrophobic (PEEK or PTFE tubings, Upchurch Scientific, INNOSEP Company, Zheng-Zhou, China) inner walls, gas-tight syringes, syringe pumps (PHD 2000, Harvard Apparatus), tubings (Polytetrafluoroethylene, Fisher Scientific Bioblock), T-junctions (P-728-01, Upchurch Scientific), line junction, and UV light (Lightning-cure LC8, Hamamatsu), as shown in Fig. 1. Two coaxial capillaries were inserted inside T-junction 1 along its main axis. The coaxial capillaries’ tips exit the T-junction 2 at the center of an outlet PTFE tubing.

Fig. 1

There are two types of disperse phase and one type of continuous phase which are mutually immiscible. Simethicone was used in continuous phase, and one of the disperse phase termed as the shell phase used TPGDA (Aldrich), 96 wt%: 1-hydroxycyclohexyl phenyl ketone (HCPK) as photoinitiator (Aldrich), 4 wt% [11]. The encapsulated phase used the core of the droplets consisted of sodium iodide solution (8 wt%) containing radioactive ¹²⁵I (5 mCi). The mass densities of titanium, TPGDA, and dry air were 4.50, 1.03, and 1.20 × 10⁻³g/cm³, respectively, and air composition by mass of nitrogen, oxygen, argon, and carbon, was 75.520, 23.176, 1.288, and 0.016%, respectively, were used for computational purposes [12]. Water consisted of two parts hydrogen, one part oxygen, with a mass density of 0.998 g/cm³.

An optical microscope (XSP-30E, Shanghai Halibut Instrument Limited Company, China) equipped with a charge-coupled device (CCD) camera (uEye, UI-2220SE, IDS) was used to observe the formation of the core-shell structure droplets directly. The camera captures up to 52 ftps at a full resolution of 768 pixels × 576 pixels. Then during the flowing in the tube, droplets were polymerized under the explosion of UV irradiation by means of an UV source operating at λ = 365 nm online. The overall diameter and the core diameter of the ¹²⁵I seed were measured with an image analysis software (uEye Cockpit) operating the CCD camera. The dosimetry parameters of the new source were calculated by Version 5 of Monte Carlo N-Particle Transport (MCNP5) code. The results of the tally were fed to a plotting software, SURFER 11.0, used to construct the isodose curves against dose values calculated above.

For the preparation of ¹²⁵I seeds, core-shell structure droplets are produced first. A flexible coaxial capillaries microfluidic device was designed by us to produce core-shell droplets consisted of liquid core surrounded by a layer of TPGDA. The main part or element of the liquid core was sodium iodide solution containing radioactive ¹²⁵I. As shown in Fig. 2b, sodium iodide solution and TPGDA were delivered by the inner capillary and the outside capillary, respectively, with the inner one was fixed inside of the outside one. Q₁, Q₂, and Q₃ were the flow rates of continuous phase, shell phase, and core phase, respectively. The reaction between surface tension and shear force in the shell phase caused the formation of core droplets [13], at the tip of the inner capillary. Subsequently, shell phase generated at the end of the outside capillary due to the shear force of continuous phase happened between the outside capillary and the PTFE tubing. As a result, the core phase was encapsulated in the shell material, and the core-shell structure droplet was fabricated. Then, the TPGDA shell was polymerized downstream by photo-induced polymerization to form the seeds with a ¹²⁵I core-PTPGDA shell structure. Polyacrylate possesses an as-desired radiation resistance. Researches of poly(tert-butyl acrylate) (PTBA) radiation resistance indicated that the thickness of irradiation-damaged polyacrylate file was about 0.1 Å after 1 minute irradiation under 9.659 keV photons at a flux density of 0.4 × 10⁹ photons/(s·mm²) [14]. For a 0.5 mCi as-prepared ¹²⁵I seed, it was estimated that the thickness of irradiation-damaged PTPGDA shell would be <1 μm after a year, which assured that TPGDA was a competent material for the shell of the ¹²⁵I seed.

Fig. 2

The droplet formation from two continuous flowing liquids was achieved through either dripping or jetting mode [15], and the key parameters for mastering the droplet formation mode are the flow rates of the continuous and dispersed phase. In this work, by adjusting the fluid flow rate, the droplets were formed in the dripping mode which ensured the formation of highly monodisperse core-shell structure droplets [16]. The formation of the core-shell structure droplet and the as-prepared ¹²⁵I seeds with uniform core size and overall seed size were shown in Fig. 2.

For the treatment of different tumour sites, a source with various radiation doses for brachytherapy is required. The dose delivered by as-prepared ¹²⁵I brachytherapy source is controlled by the dosage of ¹²⁵I in each ¹²⁵I seed and the number of the seeds sealed in the titanium capsule. Because of the limitation on the length of the source in practical utilization, controlling the content of ¹²⁵I in each seed precisely to meet the radiation dose requirements has to be approached realistically. The ¹²⁵I content of each seed is related to the ¹²⁵I concentration in feed solution of the core and the core size of as-designed seed. By keeping size of the core droplet constant, it is easy to understand that the proportional increase of radiation dose rate of the seeds with the increase of the ¹²⁵I concentration in core feed solution. While keeping the concentration of ¹²⁵I in core feed solution fixed, the content of ¹²⁵I in each seed can be controlled by changing the size of core droplet, and increases as the core size increases. In our previous work, the control of overall core-shell structure’s droplet size, core size, and shell size were studied in detail by adjusting the flow rates of different phase in the coaxial capillary microfluidic device and the relationship between core size and flow rate of the related fluids can be described by the formula: (1)rcore=Q3Q2+Q3ε×rseed{r_{{\rm{core}}}} = \root \varepsilon \of {{{{Q_3}} \over {{Q_2} + {Q_3}}}} \times {r_{{\rm{seed}}}} where r_core and r_seed, are the diameter of the core droplet and the seed, respectively, whereas Q₂ and Q³ are the flow rate of the shell phase and the core phase, respectively. In the experiments, the precise control of core size can be achieved by adjusting flow rate of core and shell phase. Also, the size-designated overall core-shell structure droplet can also be obtained by changing the flow rate of core, shell, and continuous phase, respectively. The shell phase of the core-shell structure droplets were subsequently polymerized while flowing in the tubing under ultraviolet irradiation, and ¹²⁵I seeds were fabricated finally. The as-prepared seeds with the same overall size but with different core sizes were shown in Fig. 3 and were of narrow size distribution of the cores and seeds, which were beneficial in keeping the same dose rate of the seeds. If a reduction in PTPGDA shell thickness by less than 10% is considered and still to ensure that the ¹²⁵I-containing solution was not leaked during a half-life period of ¹²⁵I (about 60 days), a single ¹²⁵I seed with the maximum radioactivity 49 mCi, with core diameter of 500 μm and shell thickness of 50 μm, calculated based on the radiation resistance of acrylate was used.

Fig. 3

A structured source with a titanium capsule size of 800 μm outer diameter, 700 μm inner diameter, and 4 mm length was used in clinical medicine conventionally and the capsules filled with six as-prepared seeds of 600 μm diameter (shown in Fig. 4) were used for an evaluation of its dose distribution, and the dosimetry parameters of the source were calculated by MCNP5 code [17]. The simulation considered a spherical water phantom and the dose values in water were calculated at radial distances of 0–2 cm. The DLC-189 cross section was used for transport calculations and the Integrated TIGER Series (ITS) three-dimensional (3D) code system was used for coupled photon–electron transport. In order to determine relevant dose distribution accurately, different size of voxel computational cells were placed in select positions. The total number of photons transported per calculation run was 10⁹ with statistical uncertainties below 3.6%. A plotting software, SURFER 11.0 was used to construct the isodose curves.

Fig. 4

If the core and shell of each seed have concentricity, a ¹²⁵I brachytherapy source containing six seeds that lined the capsule appears to have an ellipsoid-like uniform dose distribution, as shown in Fig. 4a. In fact, it is difficult to synthesize a homocentric seed in the microfluidic system since the core droplet is free to move in the double droplet before polymerization of the seed shell, which will induce nonuniform distribution of the dose.

The core center usually deviates from the seed center in microfluidic fabrication of the seed with a core-shell structure. The maximum deviation of the core center from shell center, in that the core tangent to surface of the seed, was modelled for exploring the influence of the core position on dose distribution of the source. The position of six cores was described by setting x, y coordinates in xy plane. Obviously, the more deviation of the core center from the seed center is, the less uniform dose distribution of the source exists. For a size- and position-confined seed in the capsule, the increase in the core size decreases the deviation of core center from its ideal center and more uniform dose distribution of the ¹²⁵I brachytherapy source was produced. The model of the core orientation in the source was randomly selected, as shown in Fig. 4b–d, to illustrate the influences of the core position on the dose distribution of the ¹²⁵I brachytherapy source.

The isodose curves of cores with diameter of 150, 300, 400, and 500 μm constructed by SURFER 11.0 were shown in Fig. 5, respectively, and found that when the diameter of core was 150 μm, the dose distribution had a less uniformity within ±5 mm in x axis, but beyond ±5 mm, the dose distribution was uniform. When core size was increased from 150 μm to 400 μm, the uniform dose distribution within ±5 mm was improved, but was still nonuniform. But with core size 500 μm, the dose distribution of the ¹²⁵I brachytherapy source was close to the ideal model.

Fig. 5

By keeping the core diameter of 500 μm, the dose distribution of other modelled sources with a different orientation of the cores was also computed, as shown in Fig. 6. The results indicated that the dose distribution of the ¹²⁵I brachytherapy sources with various orientations of the cores exhibited negligible differences or variations and so we can conclude that all of them were uniform even within ±5 mm in x axis and y axis. Therefore, the ¹²⁵I seeds with a core size of 500 μm and seed size of 600 μm were designed and fabricated. Additionally, an inactive dummy ¹²⁵I brachytherapy source was fabricated by encapsulating six as-prepared ¹²⁵I seeds in a glass tube with the size of 600 μm/850 μm inside/outside diameter (ID/OD), as shown in Fig. 7.

Fig. 6

Fig. 7