Size-Sorted Superheated Nanodroplets for Dosimetry and Range Verification of Carbon-Ion Radiotherapy

Abstract

Nanodroplets have demonstrated potential for the range detection of hadron radiotherapies. Our formulation uses superheated perfluorobutane (C4F10) stabilized by a poly(vinyl-alcohol) shell. High-LET (linear energy transfer) particles vaporize the nanodroplets into echogenic microbubbles. Tailored ultrasound imaging translates the generated echo-contrast into a dose distribution map, enabling beam range retrieval. This work evaluates the response of size-sorted nanodroplets to carbon-ion radiation. We studied how thesize of nanodroplets affects their sensitivity at various beam-doses and energies, as a function of concentration and shell cross-linking. First, we show the physicochemical characterization of size-isolated nanodroplets by differential centrifugation. Then, we report on the irradiations of the nanodroplet samples in tissue-mimicking phantoms. We compared the response of large (≈900 nm) and small (≈400 nm) nanodroplets to different carbon-ions energies and evaluated their dose linearity and concentration detection thresholds by ultrasound imaging. Additionally, we verified the beam range detection accuracy for the nanodroplets samples. All nanodroplets exhibited sensitivity to carbon-ions with high range verification precision. However, smaller nanodroplets required a higher concentration sensitivity threshold. The vaporization yield depends on the carbon-ions energy and dose, which are both related to particle count/spot. These findings confirm the potential of nanodroplets for range detection, with performance depending on nanodroplets' properties and beam parameters.

Keywords: carbon-ions radiotherapy; dosimetry; nanodroplets; range verification; ultrasound imaging.

Figure 1

Schematic illustration of differential centrifugation steps for the size sorting of PVA/PFB NDs. The inset picture is a photograph of size-sorted NDs: large (left vial), medium (middle vial), and small (right vial).

Figure 2

Bright-field microscopy images (objective 40×) of Neubauer counting chambers filled with size-sorted PVA/PFB nanodroplet samples obtained by selective centrifugation and prepared with two sonication durations: (a) 15 min of sonication (dilution ×50) and (b) 25 min of sonication (dilution ×100). (c,d) DLS intensity-weighted size distributions of PVA/PFB ND separated populations by selective centrifugation corresponding to (a,b), respectively. Scale bars: 10 μm.

Figure 3

Concentration variation over time of size-sorted PVA/PFB NDs, i.e., L-NDs (blue), M-NDs (red), and S-NDs (black) produced by (a) 15 min sonication and (b) 25 min sonication. (c,d) Corresponding mean diameter variation over time of the size-sorted NDs produced by 15 min and 25 min sonication, respectively. The lines are a guide for the eye.

Figure 4

ADV study of size-sorted PVA/PFB NDs assessed by bright-field microscopy (long-distance objective 40×) in ibidi µ-channel slides. We show the ND images upon channel loading (pre-incubation), after thermalization at 37 °C, and after acoustic activation using 1 MHz US transducer (Sonidel, 2 W/cm², 30 s). On the right side we show the size distributions of the generated microbubbles upon phase-transition, corresponding to each size-sorted ND sample. The ND samples were produced after 25 min sonication. Scale bars: 10 μm.

Figure 5

Comparison of the size-sorted ND (25 min sonication) response to C-ions (312 MeV/u, 180 mm range) at 37 °C after exposure to doses between 0.1 and 4 Gy: (a) depth-resolved US images (7.5 MHz, MI = 0.1) of independent phantoms of NDs dispersed in PAM (8.4 × 10⁵ NDs/mL) post-irradiation at the different tested doses (the yellow arrows indicate the C-ions beam’ entrance side). (b,c) Average contrast gray-value profiles from the vaporization of L-NDs and S-NDs, respectively, at the various doses (number of frames/phantom = 3). (d,e) Evaluation of the peaks’ integrals from the average grayscale profiles of L-NDs and S-NDs, respectively, as a function of C-ion dose. The inset shows the linear regression fit up to 2 Gy.

Figure 6

(a) Average contrast gray-value vaporization profiles extracted from US images (inset figures, MI = 0.1, 7.5 MHz) of S-NDs (obtained after 15 min sonication) post1Gy C-ion irradiation with different ranges between 30 and 180 mm at 37 °C (the pre-irradiation background profile was subtracted). The ND concentration in the phantoms is 10⁶ NDs/mL. (b) Corresponding evaluation of both vaporization peaks integrals and FWHM from the average grayscale profiles of the NDs as a function of the C-ion range (the curves are fitted with an exponential function). (c) Summary plot of measurements of echo-contrast peak integrals of S-NDs (15 min sonication, C ≈ 10⁶ NDs/mL) irradiated under different doses/ranges as a function of C-ion particle count/spot (the red line is a logistic fit).

Figure 7

(a) Comparison of US images (MI = 0.1; 7.5 MHz) of L-NDs (C ≈ 10⁶ NDs/mL), obtained with sonication durations of 15 min and 25 min, after irradiation with 1 Gy C-ions at 30 mm and 50 mm ranges. (b,c) Comparison of the gray-value vaporization profiles of L-NDs obtained at the different sonication durations corresponding to the exposure to C-ions with ranges of 30 mm and 50 mm, respectively. The pre-irradiation background was subtracted from the profiles.

Figure 8

(a,b) US image (MI 0.1, 7.5 MHz) of S-NDs and L-NDs post 1 Gy SOBP C-ions (160–180 mm) and (c) corresponding average gray contrast-value profile corresponding to generated MBs. Both NDs samples were produced with 15 min of sonication. The concentration of L-NDs and S-NDs in the phantoms is C ≈ 10⁶ NDs/mL. The yellow arrows indicate the beam entrance side.

Figure 9

(a,b) US images of L-NDs (25 min, C ≈ 10⁶ NDs/mL) pre and post2Gy multi-energy C-ions irradiation(i.e., 7 energies from 126 to 174 mm with energy step of 8 mm). The yellow arrows indicate the beam entrance side; (c) top-view photograph of the irradiated phantom and (d) corresponding contrast gray-value vaporization profile of NDs (the red arrow in d is an imaging artifact caused by the combination of images to cover the full length of the phantom). (e) US image of S-NDs (15 min, C ≈ 10⁶ NDs/mL) post 5 Gy single spot C-ion irradiation with an energy of 399 MeV/u (the US transducer is perpendicular to the beam direction); (f) corresponding contrast gray-value profile (the red line is a Gaussian fit) and (g) 3D US map reconstructed from a full scan of the spot with two projections (i.e., perform scanning with the transducer perpendicular and parallel to the beam direction). The C-ion total count in the spot is 8.2 × 10⁶.Pre-irradiation profiles were subtracted.