Hepatocellular carcinoma (HCC) is one of the most common cancers in the world and poses a major threat to human health [1, 2]. Current therapeutic strategies for HCC in the clinics mainly comprise surgical resection, liver transplantation, radiofrequency ablation, and transarterial embolization [3-7]. Among those strategies, transarterial embolization as an attractive treatment method, including transarterial chemoembolization and transarterial radioembolization (TARE), kills the tumor cells by blocking the hepatic artery [8-11]. TARE is a local radionuclide therapy that relies on radioactive materials, like gel, microparticles, and microspheres, to block the artery, so as to maximize the radiation dose in the tumor and minimize the damage to normal liver tissue. TARE is suitable for treating patients who suffer from unresectable liver cancer, or fail from other therapies [12-14].

Radioactive microspheres have been demonstrated to be an ideal carrier for TARE [15, 16]. To date, three types of radioactive microspheres are clinically available, including ⁹⁰Y-labeled glass microspheres (TheraSphere), ⁹⁰Y-labeled resin microspheres (SIR-Spheres), and ¹⁶⁶Ho-labeled poly(l-lactic acid) (PLLA) microspheres (QuiremSpheres). However, all these types of microspheres suffer from some shortcomings. For instance, ⁹⁰Y-glass microspheres and ¹⁶⁶Ho-microspheres are produced by neutron activation, which not only requires a nuclear reactor but also produces simultaneously other long half-life radioisotopes unwanted. Although the production of ⁹⁰Y-resin microspheres does not have the above problems, the radiostability of the microspheres needs to be improved. In addition, ⁹⁰Y is a pure beta emitter. It is therefore not suitable for biomedical imaging, let alone the evaluation of possible leakage of the loaded ⁹⁰Y in vivo [17-21]. In order to overcome these limitations, a variety of different radioactive microspheres have been developed, such as ⁹⁰Y polymeric microspheres, ¹³¹I polymeric microspheres, ¹⁸⁸Re-human serum albumin microspheres, and ¹⁸⁸Re polymeric microspheres [22-28]. However, ¹³¹I emits high energy gamma ray (364 keV), which may cause damage to healthy organs, especially the lung. The half-life of ¹⁸⁸Re is as short as 16.9 h, which is inconvenient for delivery from the production site to the customer. Moreover, the initial injected activity of ¹⁸⁸Re has to be approximately 2.8 times higher than that of ⁹⁰Y to achieve a comparable dose, which requires more shielding precaution for protecting the physicians.

Apart from the carrier used in radioembolization, the other important factor that affects the therapeutic efficiency is the radionuclide. Among many radionuclides, ¹⁷⁷Lu is a very attractive one. Because it emits medium-energy beta particles (490 keV) and low-energy gamma rays (113 keV (3%), 210 keV (11%)), which provides an opportunity for observing the positioning of ¹⁷⁷Lu labeled microspheres with single-photon emission computed tomography (SPECT) imaging apart from radioisotope therapy. In addition, the proper half-life of ¹⁷⁷Lu, i.e., 6.7 d, makes it suitable for transportation and clinical application [29, 30]. Therefore, ¹⁷⁷Lu has been used in peptide radioligand therapy for prostate cancers and peptide receptor radionuclide therapy for neuroendocrine tumors [31-34]. More importantly, there exist several reactors in the world that can commercially supply ¹⁷⁷Lu.

In this study, we developed a facile and efficient method for preparing ¹⁷⁷Lu-microspheres by directly precipitating ¹⁷⁷Lu on commercial silica microspheres for HCC radioisotope therapy and imaging (Scheme 1). Silica microspheres are a kind of material that is easy to prepare, has a controllable morphology and good biocompatibility [35, 36]. The obtained ¹⁷⁷Lu-microspheres showed both controllable specific activities and high radiostability. HepG2 subcutaneous tumor-bearing mice were used for in vivo biodistribution and anti-tumor inhibition evaluation. In vivo biodistribution monitored by microSPECT indicated that ¹⁷⁷Lu-microspheres were retained in the tumor even 32 d after intratumoral injection. Most importantly, the above ¹⁷⁷Lu-microspheres presented a remarkable anti-tumor inhibition efficiency.

¹⁷⁷Lu-microspheres were prepared through a precipitation method by using silica microspheres as matrix. K₃PO₄ and KOH were used to adjust the pH of the reaction medium as the precipitation reactions of ¹⁷⁷Lu³⁺ are generally pH-dependent. As shown in Fig. 1A, the radiolabeling efficiency increases dramatically against pH and reaches a plateau around 96.8% ± 0.5% at pH 12. In addition, the effect of reaction time on the radiolabeling efficiency was also studied. According to the results given in Fig. 1B, the radiolabeling efficiency higher than 96% can be achieved within 1 min, nearly independent of the reaction time. The following studies suggested that the specific activity, calculated by dividing the radioactivity of ¹⁷⁷Lu by the number of microspheres, could be varied from 1 Bq/microsphere to 1 × 10⁵ Bq/microsphere through the approach mentioned above. High specific activity is favorable for avoiding arterial stasis caused by excessive microspheres, as less number of microspheres is required for achieving the same radiation dose [37, 38]. To characterize the morphology, size and size distribution of the microspheres after ¹⁷⁷Lu labeling, ¹⁷⁵LuCl₃ was used instead of ¹⁷⁷LuCl₃ to prepare non-radioactive microspheres (denoted as Lu-microspheres) following the same preparative procedures for ¹⁷⁷Lu-microspheres. The average size of the Lu-microspheres determined by optical microscopy is of 20.7 ± 0.7 µm and the polydisperse index is of 0.034, as shown in Fig. 1C and Fig. S1A (Supporting information). The scanning electron microscopy (SEM) image given in Fig. 1D reveals that the Lu-microspheres are perfectly spherical in shape. In comparison with the mother silica microspheres, the Lu-microspheres present nearly unchanged shape, size and size distribution, suggesting that the coprecipitation approach did not alter the morphology of the mother silica microspheres at all (Figs. S1B-D in Supporting information). The radiostability of ¹⁷⁷Lu-microspheres in saline and 10% FBS was tested and the free ¹⁷⁷Lu in saline and 10% FBS was measured to be less than 2% and 20% (Fig. 1E), respectively, indicating that ¹⁷⁷Lu-microspheres are very stable for further in vitro and in vivo experiments. The transformation of ¹⁷⁷Lu to ¹⁷⁷Hf is accompanied by emission of gamma rays of 113 keV (3%) and 210 keV (11%) which are suitable for SPECT imaging. To verify the SPECT imaging capacity of the ¹⁷⁷Lu-microspheres, tube phantom imaging was carried out. The results given in Fig. 1F and Fig. S2 (Supporting information) reveal that the signal intensity extracted from SPECT images increases linearly against the radioactivity of ¹⁷⁷Lu-microspheres, suggesting that the ¹⁷⁷Lu-microspheres are potentially suitable for SPECT imaging of tumors in vivo. Most importantly, the SPECT signal offers a non-invasive way to track the ¹⁷⁷Lu-microspheres in vivo.

Before the in vivo study, the potential cytotoxicity of the mother silica microspheres and ¹⁷⁷Lu-microspheres to HepG2 liver cancer cells was evaluated by MTT assay. According to the results given in Fig. 2A, no obvious cytotoxicity was observed even at silica microspheres concentration of 500 µg/mL. However, at the same concentration of silica microspheres, 50 µCi radioactivity can induce significant cytotoxicity to HepG2, e.g., the cell viability is decreased to 42.01% ± 8.02%, as shown in Fig. 2B. Those results could be explained by more energy deposited on cells, due to ¹⁷⁷Lu-microspheres close to the cells. In addition, the cytotoxicity of silica microspheres in normal cells was further tested. No obvious cytotoxicity was observed even at silica microspheres concentration of 500 µg/mL in 3T3 mouse embryonic fibroblast cells as shown in Fig. S3 (Supporting information).

The emission of electron of 490 keV during the transformation of ¹⁷⁷Lu to ¹⁷⁷Hf can induce DNA damage [39]. To evaluate this effect, an immunofluorescence assay was carried out to analyze the DNA damage. It was confirmed that ¹⁷⁷Lu-microspheres give rise to obvious DNA breakages contrasting to the mother silica microspheres (Fig. 2C). Further quantifying the immunofluorescence imaging results reveals that the fluorescence intensity induced by ¹⁷⁷Lu-microspheres is increased by a factor of 31 in comparison with that induced by silica microspheres, indicating that ¹⁷⁷Lu-microspheres possess a remarkable potential for treating HepG2 liver cancer (Fig. 2D).

The in vivo embolization of ¹⁷⁷Lu-microspheres was evaluated through a rabbit ear model. After 8 d embolization, the tip of rabbit right ear was avascular necrosis due to the obstruction of blood flow as shown in Fig. S4 (Supporting information), indicating that ¹⁷⁷Lu-microspheres could achieve good embolization.

Prior to the anti-tumor treatment, the imaging performance of ¹⁷⁷Lu-microspheres was evaluated on HepG2 cell tumor-bearing BALb/c nude mice by using free ¹⁷⁷Lu³⁺ as control. After intratumoral injection, SPECT scans were conducted at different time points postinjection. It can be observed that the signal of ¹⁷⁷Lu-microspheres remains visualizable for 32 d postinjection (Fig. 3A). More importantly, no radioactivity was found in other organs or tissues, suggesting that ¹⁷⁷Lu-microspheres were very stable in vivo. According to the quantification analysis of the SPECT images, the radioactivity of ¹⁷⁷Lu-microspheres in tumor decreases against time, perfectly consistent with the theoretical physical half-life attenuation of ¹⁷⁷Lu (Fig. 3B), supporting that the ¹⁷⁷Lu-microspheres in the tumor do not release the ¹⁷⁷Lu payload to other organs. In contrast, free ¹⁷⁷Lu gives rise to visualizable signal only for 12 d postinjection, due to the complexation between ¹⁷⁷Lu³⁺ and protein [40]. On the one hand, the tumor became too big (1500 mm³) after that to allow the continuation of the experiment and on the other hand ¹⁷⁷Lu was found to diffuse obviously from the tumor site to whole body and accumulate particularly in bone. In consequence, the radiation dose is dramatically decreased as shown in Fig. 3B.

To accurately trace the distribution of ¹⁷⁷Lu, the radioactivity of organs and tissues extracted after the imaging study were measured by γ counter. The results given in Fig. 3C confirm that almost no ¹⁷⁷Lu ion is released by ¹⁷⁷Lu-microspheres, while apart from tumor (3.1% ± 0.6% ID/g), radioactivity of ¹⁷⁷Lu can also be detected from other organs and tissues such as bone (2.8% ± 0.05% ID/g) if ¹⁷⁷Lu is delivered in the form of LuCl₃. It was also observed that it took just a few hours for bone to show evident signal post-administration of free ¹⁷⁷Lu (Fig. S5 in Supporting information), eventually leading to significant signal at the end. This is because as a lanthanide element lutetium has a strong osteophilicity in the body. All these results suggest that ¹⁷⁷Lu-microspheres hold a great potential as a radiopharmaceutical for liver cancer treatment.

The tumor-therapeutic effect of the ¹⁷⁷Lu-microspheres was evaluated through tumor inhibition experiments on HepG2 cell tumor-bearing BALb/c nude mice. The tumor-bearing mice were randomly divided into four groups and received saline, mother silica microspheres, ¹⁷⁷Lu-microspheres with different doses, i.e., 100 µCi and 1 mCi, respectively. The cumulative tumor radiation of 1 mCi and 0.1 mCi groups amounted to 551.3 Gy/g and 55.1 Gy/g, respectively, within the first 14 d postinjection. Notably, the ¹⁷⁷Lu-microspheres can significantly inhibit the tumor growth (Fig. 4A), particularly the 1 mCi group. Although the tumors of the 1 mCi treatment group didn't completely vanish, a better treatment efficacy can still be expected because radioembolization will surely give rise to a better distribution of than ¹⁷⁷Lu-microspheres than that achieved by intratumoral administration.

The volumes of the tumors were recorded every 2 days with a vernier caliper, and the body weights were also monitored. As shown in Fig. 4B, the treatment with ¹⁷⁷Lu-microspheres does not lead to obvious loss of body weight compared to other reference groups, demonstrating that the ¹⁷⁷Lu-microspheres have no obvious side effects. The photos of representative mice from each group are shown in Fig. 4C.

The physiological conditions of the mice look very good. As the tumor volume of control group reached 1500 mm³, the mice were sacrificed according to ethical rules. The tumors collected from different groups are shown in Fig. 4D for comparison.

The tumor and main organs including heart, liver, spleen, lung, and kidney of the mice receiving different treatments were sliced and then subjected to hematoxylin and eosin (H & E) and TUNEL staining. As shown in Fig. 5, the treatment with ¹⁷⁷Lu-microspheres can apparently give rise to remarkable necrosis and apoptosis in comparison with the control group and silica microspheres group. More importantly, the H & E staining reveals that nearly no damage was created by the ¹⁷⁷Lu-microspheres groups to major organs as given in Fig. 6.

In summary, we have developed a facile and effective approach for preparing ¹⁷⁷Lu-microspheres capable of inhibiting the tumor growth apart from being visualizable with SPECT. Through simple precipitation method, the radiolabeling efficiency as high as 96.8% ± 0.5% has been successfully achieved and the specific activity of the resulting ¹⁷⁷Lu-microspheres with outstanding radiostability can be controlled over a broad range from 1 Bq to 1 × 10⁵ Bq/microsphere. Over 30-day observation demonstrates that the ¹⁷⁷Lu labeling is super robust in vivo. The tumor therapeutic studies reveal that the ¹⁷⁷Lu-microspheres can effectively inhibit the growth of HCC, demonstrated on the subcutaneous mice tumor model, and induce no obvious side effects. Therefore, ¹⁷⁷Lu-microspheres hold great potential as radioembolic microspheres for precise radioisotope therapy of HCC.

Ethical statement

All animal experiments were approved by the Animal Care and Use Committee of Soochow University, and all protocols of animal studies conformed to the Guide for the Care and Use of Laboratory Animals of Soochow University, China.

Declaration of competing interest

The authors declare no competing financial interest.

Acknowledgments

The authors thank the financial support from the National Natural Science Found of China (Nos. 81720108024, 21976128), Priority Academic Program Development of Jiangsu Higher Education Institutions (PAPD). G. Wang thanks the support from the Natural Science Foundation of Jiangsu Province (No. BK20200100).

Supplementary materials

Supplementary material associated with this article can be found, in the online version, at doi:10.1016/j.cclet.2022.01.007 .