Breast cancer is the most common cancer globally. Currently, breast cancer affects approximately 12% of women globally (1). While the outcomes of localized primary breast cancer can be successfully eradicated by surgery with promising survival, the outcomes of metastatic breast cancer are abysmal. Liver metastases comprise half of all breast cancer malignancies, carrying with them an inauspicious prognosis and a scant 5-year survival of 8.5% (2). Medical treatment of metastatic breast cancer is directed by tumor subtype, such as hormonal therapy for estrogen receptor (ER)–positive subtypes and trastuzumab for Human epidermal growth factor receptor 2 HER2-positive subtypes. For hepatic metastasis, surgery, external beam radiation therapy (EBRT), and logoregional therapy such as chemoembolization and thermoablation are also effective in reducing tumor burden and prolonging survival (3). Despite a variety of treatment options, it is difficult to treat patients with large metastatic breast cancer to the liver recalcitrant to medical therapy and cannot tolerate surgery or EBRT. Selective internal radiation therapy (SIRT) with yttrium-90 (Y-90) has emerged as an effective treatment for unresectable hepatocellular carcinoma, cholangiocarcinoma, and secondary liver tumors such as colorectal cancer and uveal melanoma metastasis (4–7). In the last few decades, several single-center retrospective cohort studies reported its use in metastatic breast cancer to the liver (6–14). The aim of the present study is to investigate the safety and efficacy of SIRT in treating metastatic breast cancer through meta-analysis.

The present meta-analysis reviewed the available evidence and suggested that SIRT was feasible in treating patients with breast cancer with hepatic metastasis recalcitrant to other therapies with an overall post-embolization MST of 9.8 months. Patients with <25% and lack of extrahepatic disease showed a better response.

SIRT implements resin or glass microspheres embedded with radioactive isotope yttrium into tumors. Each bead exerts radiotherapy directly to adjacent tissues with approximately 2.5-mm penetration (33). Meanwhile, these beads also deprive tumors from arterial blood supply similar to TACE and bland Transarterial Embolization (TAE), leading to tissue ischemia and tumor necrosis. Although more than 70% blood supply of normal liver parenchyma derives from the portal venous system, hepatic malignancy is mainly supported by arteries. Tumor localization using transarterial microcatheters ensures SIRT’s selective delivery, spares normal tissue, and preserves liver functional reserve. Using SIRT segmentecotomy, >190 Gy of radiation dosage can be selectively delivered to a tumor-containing hepatic segment to achieve complete necrosis (34, 35). Because of its treatment precision, SIRT has been implemented for patients with Hepatocellular Carcinoma (HCC) who cannot tolerate sorafenib’s adverse effects. While initially considered as a palliative treatment for advanced stage HCC, SIRT has gradually been recognized for its curative role and is now an option for early and very early stage HCC according to the most recent Barcelona Clinic Liver Cancer guideline (36). The present study echoes evidence of SIRT on HCC by showing that SIRT can be a feasible option as a salvage treatment for metastatic breast cancer.

Available evidence of SIRT in breast cancer with hepatic metastasis mainly focused on tumors that are unresectable, recalcitrant to systemic treatment, and/or high liver disease burden. The 5-year survival rate of patients with metastatic breast cancer to the liver was 10% after systemic therapy (37), whereas the 5-year survival rate was 54% for patients with resectable disease. For unresectable breast cancer liver metastases, percutaneous ablative therapy using radiofrequency, laser, and microwave ablation are also feasible, with an MST of 10.9–54 months. The 5-year survival rate was 27%–35%, whereas local tumor progression was 2.9%–9.5% (38). Nonetheless, the seemingly more favorable outcomes associated with surgical and ablative treatment could be attributed to milder tumor burden compared with patients subjected to embolotherapy, which are often too large for resection and ablation. For breast cancer with heavy liver metastasis burden, transarterial chemoembolization, a treatment for BCLC stage B liver cancer and large colorectal metastasis, has been utilized with a promising MST from 7.3 to 47 months (39). In treatment of HCC, however, SIRT has gradually gained increasing popularity in the last decade. More pieces of evidence also became available supporting its use in metastatic cancer to the liver, such as colorectal metastasis and melanoma (7, 40). With both TACE and SIRT as the available options, comparative studies showed superior survival in among patients treated with the former (41, 42). As for breast cancer with liver metastasis, only Chang et al. compared TACE and SIRT in treating liver metastatic breast cancer, suggesting a longer MST (4.9 vs. 12.9 months) and fewer adverse events of SIRT (71% vs. 44%) (9). Further randomized controlled trials are warranted to validate these findings.

Compared with SIRT, EBRT has been historically more commonly used for unresectable hepatic metastasis. For breast cancer metastasized to the liver, stereotactic body radiation therapy (SBRT) offers a 1-year OS rate of 21%–85% (43–45). The 1- and 2-year local control rates, as measured by PERCIST criteria, were 100% and 80%, respectively (44). Compared with SIRT, EBRT of the liver faces the challenges of respiratory motion, requirement of fiducial marker, increased radiation to the normal parenchyma, multiple treatment sessions, limited dose delivery in the setting of large tumors, etc. (46, 47). While the comparative studies between EBRT and SIRT were lacking in the setting of breast cancer metastasis to the liver, a previous study on cholangiocarcinoma showed a seemingly longer survival of SIRT over EBRT as the first-line therapy (MST of 36 vs. 11 months) (48), highlighting SIRT’s effectiveness in reducing hepatic tumor burden as a novel therapy compared with the more traditional EBRT approach.

On the basis of the present study, the pooled TRR and TCR were 36%–49% and 73%–97%, respectively. These findings are similar to previously reported 22%–81% and 78%–96% in the primary liver cancer such as cholangiocarcinoma (48). Radiological response is a well-known predictor of survival after SIRT of HCC and colorectal liver metastasis (49, 50). According to Saxena et al., the radiological response also correlates with survival of patients with breast cancer with liver metastasis after SIRT (14). Among a variety of imaging criteria, RECIST and WHO criteria were developed first to characterize tumor response toward treatment by measuring uni- and bi-dimensional measurements in the liver cancer. Because anti-cancer effects can manifest as necrosis while maintaining a stable size, mRECIST is established to take this aspect into consideration (51). PERCIST, by contrast, can further evaluate the functionality of tumor by measuring glucose uptake (52). Although all these response criteria have been implemented in evaluating SIRT of the primary and secondary liver cancers, PERCIST and mRECIST have recently gained increasing popularity due to the accurate predictability of survival outcomes in hepatocellular carcinoma (53, 54). For metastatic breast cancer to the liver, the ideal radiological response criteria for survival prediction remain to be determined by the future comparative studies.

Whereas the included individual studies are heterogeneous in terms of patient population and disease burden, the present meta-analysis demonstrated that that the lower tumor burden and the absence of extrahepatic disease were associated with improved survival. Other reported factors include adjuvant chemotherapy, eastern cooperative oncology group performance status (ECOG) status, tumor vascularity, estrogen receptor status, baseline serum bilirubin and transaminase level, and [¹⁸F]Fluorodeoxyglucose (18F-FDG) standardized uptake value (13, 14, 21, 23–25). Some of these trends were consistent with the previous literature focusing on other treatment modalities in metastatic breast cancer and SIRT in HCC or colorectal cancer (55, 56). These findings advocate the early consideration of SIRT in patients with lower disease burden to achieve a more promising survival. The recent LEGACY study showed that SIRT could achieve 3-year OS rate of 87% for patients at early-stage HCC with ECOG 0-1 and unresectable tumor of up to 8 cm with 21% bridged to transplant or resection. The use of SIRT in breast cancer with limited hepatic metastasis burden is to be validated by the future studies.

In terms of safety, SIRT of breast cancer with liver metastasis is well tolerated without procedure-related death. One of the most serious side effects associated with SIRT, i.e., radiation-induced liver disease (RILD), did not occur among the included studies. On the basis of the evidence on HCC and colorectal metastasis, RILD typically has a reported incidence rate of less than 10% (57). The incidence rate of major gastrointestinal complications such as ulcer, cholecystitis, and pancreatitis were less than 3%. Careful angioanatomical planning and techniques such as coiling non-target vessels could decrease these risks. Radiation pneumonitis was not noted in the present study, which is rare nowadays due to measurement of lung-shunting fractions using dosimetry (58). Furthermore, the risk of retroperitoneal hematoma during femoral approach is significantly decreased with the use of closure devices. Alternative transradial approach has also gained popularity in SIRT, avoiding major vascular complications such as hematoma, pseudoaneurysm, and fistula formation.

The present meta-analysis should be interpreted with several caveats. First, the included studies were heterogeneous. Patients with different breast cancer subtypes, previous cancer therapy, concurrent chemotherapy, and disease burden were included. Stratification based on these variables is technically implausible with reported study-level outcomes. Second, the efficacy of SIR could be undermined because it is used as salvage treatment after patients failed multiple lines of treatments. Even patients with terminal disease and extra hepatic burden were included. The baseline survival of such patient population should be poorer than the survival rates reported in the literature, calculated from the diagnosis of hepatic breast metastasis. Third, radioembolization techniques were not specified. Whether superselective segmentectomy or whole liver Embolization was adopted would affect procedural safety profile. Furthermore, breast cancer with liver metastasis can be either hypo- or hypervascular (59). Hypervascular lesions could lead to higher radioembolization bead deposit and is more responsive toward SIRT, which was not analyzed in the present study. Last but not the least, most of the included studies were retrospective in design, categorized as level IV evidence. Future randomized controlled trial studies enrolling patients with a more homogenous baseline characteristics are warranted.

SIRT is a feasible and effective treatment for breast cancer with liver metastasis. Patients with a low liver tumor burden and the lack of extrahepatic metastasis are more likely to convey favorable survival. Tumor responded toward SIRT on radiological follow-up evaluation. No life-threatening adverse effect occurred. However, the included patient population was heterogeneous in treatment history and disease severity, limiting the ability to draw broad conclusions. Given the non-comparative nature of most studies, future prospective and multicenter randomized controlled trials are warranted to determine the comparative efficacy of SIRT versus the other treatment approaches.

The original contributions presented in the study are included in the article/ Supplementary Material . Further inquiries can be directed to the corresponding author.

CL: Manuscript writing, data collection, statistical analysis GT: Manuscript writing, data collection QS: Data collection LM: Data collection JJ: Data collection ZY: Manuscript revision, guarantor QY: Manuscript writing/revision, statistical analysis, conceptualization, guarantor. All authors contributed to the article and approved the submitted version.

This study is funded by grants from the Natural Science Foundation of Shandong Province (ZR2019MH109).

The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.

All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article, or claim that may be made by its manufacturer, is not guaranteed or endorsed by the publisher.