Stereotactic Radiofrequency Ablation of Breast Cancer Liver Metastases: Short- and Long-Term Results with Predicting Factors for Survival

Purpose To evaluate safety, local oncological control, long-term outcome and potential prognostic factors of stereotactic RFA (SRFA) for the treatment of BCLMs. Methods Between July 2003 and December 2019, 42 consecutive female patients with median age 54.0 years were treated with SRFA at our institution for 110 BCLMs in 48 ablation sessions. Median tumor size was 3.0 cm (0.8–9.0). Eighteen (42.9%) patients had extrahepatic metastasis at initial SRFA. Results Technical success rate was 100%, i.e., all coaxial needles were inserted with appropriate accuracy within 10 mm off plan and 107/110 (92.3%) BCLMs were successfully ablated at initial SRFA. Four Grade 1 (8.3%, 4/48) and one Grade 2 (2.1%, 1/48) complications occurred. No perioperative deaths occurred. Local recurrence developed in 8 of 110 tumors (7.3%). Overall survival (OS) rates of all patients at 1, 3, and 5 years from the date of the first SRFA were 84.1%, 49.3%, and 20.8% with a median OS of 32.3 months. Univariable cox regression analyses revealed age > 60 years and extrahepatic disease (without bone only metastases) as significant predictors of worse OS (p = 0.013 and 0.025, respectively). Size and number of metastases, hormone receptor status and time onset did not significantly affect OS after initial SRFA. Conclusions SRFA is a safe, minimally invasive treatment option in the management of BCLMs, especially in younger patients without advanced extrahepatic metastasis, including those with large liver tumors.


Introduction
Breast cancer is one of the most common malignancies in women and is a leading cause of mortality worldwide [1]. Approximately, 20% of breast cancer patients develop metastatic disease [2], with the lungs, liver, bone and brain being the most common sites. Breast cancer liver metastasis (BCLM) confer a poor prognosis of 4-8 months survival [3] and are found in approximately 50% of patients with metastatic disease, where 5-12% of patients have liver only metastases [3]. However, patients with negative resection margins after surgery show 5-year survival rates up to 40% [4].
Advanced breast cancer is primarily treated by systemic hormone therapy and/or chemotherapy, and despite advancements in systemic treatment, median overall survival and 5-year survival rates remain low, at 18-24 months and 27%, respectively [5].
Recent studies suggest that subgroups of breast cancer patients with oligometastatic disease benefit from additional locoregional treatment [6,7], which is defined by the 3rd ESO-ESMO (European School of Oncology-European Society for Medical Oncology) consensus guidelines as limited metastatic disease with up to five metastases that are potentially amenable for local treatment [8].
Percutaneous thermal ablation methods, such as radiofrequency ablation (RFA) and microwave ablation (MWA) have gained widespread acceptance as a minimally invasive treatment option in the management of primary and metastatic liver tumors [9][10][11][12]. Despite several studies showing promising results for RFA in BCLM [13,14], high-quality evidence is still lacking.
Stereotaxy (derived from the Greek meaning ''solid arrangement''), allows the planning of complex trajectories using three-dimensional image datasets with precise transformation into real patients using a Cartesian coordinate system [15]. Furthermore, fusion with previously acquired MR images in case of poor tumor visibility, immediate post-ablation contrast-enhanced CT fusion with the planning CT for reliable assessment of ablation results, allows more complex interventions such as large tumors and those in challenging localizations such as the hepatic dome or caudate lobe.
The purpose of the present study was to evaluate safety, local oncological control, long-term outcome and potential prognostic factors of stereotactic RFA (SRFA) for the treatment of BCLMs.

Patient Selection
The local institutional review board approved this retrospective single-center study, and all patients included gave their informed consent. Each case was reviewed and the treatment plan was approved by consensus in multidisciplinary tumor advisory board meetings.
One thousand seven consecutive patients were treated by SRFA between July 2003 and December 2019. Twentyeight patients who underwent SRFA for benign liver tumors were excluded. Forty-two consecutive patients with BCLM were treated in 48 ablation sessions and included in the study. Table 1 shows the baseline characteristics of the study group.
Exclusion criteria for SRFA comprised (i) platelet count \ 50,000/mm3 (ii) prothrombin activity \ 50% and (iii) tumor location close to (\ 10 mm) the central bile ducts. Tumor diagnosis was confirmed by multiphasic contrast MRI or CT and inconclusive cases were validated by biopsy.

Outcome Measurements
Sufficiently precise coaxial needle placement (deviation from plan \ 1 cm at each needle tip) was defined as a technical success. Technical efficacy and local recurrence rate (LR) were determined by contrast-enhanced CT or MR follow-up examinations performed at intervals of 1 month  [19]. Disease-free (DFS) and overall survival (OS) were calculated from the date of initial SRFA until the date of death due to malignancy or other causes (i.e., event), until date of relapse (DFS) or until the last follow-up visit (i.e., censoring).

Statistics
Statistical analysis was performed using IBM-SPSS v24. Data were expressed as total numbers, median and range. Differences between categorical variables were evaluated using the X 2 test and between independent continuous

Periprocedural Complications
According to the CIRSE Classification System for Complications, four Grade 1 (8.3%, 4/48) and one Grade 2 (2.1%, 1/48) complications occurred. Four patients developed arterial bleeding from subcapsular liver vessels, managed by transarterial coil embolization in the same anesthetic session. One major pleural effusion required treatment with a chest tube. Median hospital stay after SRFA was 4.5 days, ranging from 2-39 days. No perioperative deaths occurred.

Local Tumor Control and Distant Tumor Progression
Technical success rate was 100%, i.e., all coaxial needles were inserted with appropriate accuracy within 10 mm off plan. 107/110 liver metastases were successfully ablated at initial SRFA (97.3% primary technical efficacy rate), whereby 1 of 3 tumors were successfully treated in a second session, resulting a secondary technical efficacy rate of 98.2%. Local recurrence (LR) developed in 8 of 110 tumors (7.3%) after a median imaging follow-up of 10.9 months (range 1.4-112 months). Details for insufficient local control are presented in Table 2. An overview of success rates is provided in Table 3.
There was no significant difference in LR for lesions [ 3 cm and \ 3 cm with LR rates of 11.5% (3/26) and 6.0%   Table 4.
Multivariable cox regression analyses showed no independent prognostic factors for OS. Disease-free survival (DFS) for all patients at 1, 3, and 5 years from the date of the first SRFA were 45.3%, 22.3%, and 15.9% with a median OS of 10.5 months (95% CI 6.8-25.0).

Discussion
The results of the present study suggest that patients with breast cancer liver metastases benefit from stereotactic radiofrequency ablation (SRFA). More specifically, we found a median OS of 32.2 months from the date of BCLM treatment, which is considerably higher than no treatment, at 3-15 months [20,21]. In contrast to the substantial evidence behind resection for colorectal liver metastasis, data regarding BCLM are limited. In a systematic review of 33 papers, Fairhurst et al. [22] reported a median OS of 35.1 months and a 5-year survival rate of 33% after BCLM resection, which is closely aligned with the data of the present study. Newer studies such as Ruiz et al. [23] reported better OS when combining resection with systemic treatment, with a median OS of up to 82 months for liver confined metastases. Regarding survival after conventional RFA in patients with BCLM, several authors showed promising results. The reported median OS rates of several studies range from 26-29.9 months with OS rates at 1, 3 and 5 year of 68-90%, 25-44% and 11-27% [14,24,25]. These studies included mainly small tumors up to 3 cm due to the limitations of conventional targeting techniques. Despite the fact that the majority (62%) of patients in the present study had multiple liver metastases with a median size of 3 cm (up to 9 cm), our median OS of 32.2 months, with 1-, 3-and 5-year OS rates of 84.1%, 49.3% and 20.8% compare favorably to the literature regarding conventional RFA.
Patient selection, especially in the context of advanced breast cancer patients, is a crucial step towards improving outcomes, which could be improved by identifying prognostic factors associated with better survival. Positive hormone receptor status has been reported as a positive predictive factor for survival after hepatic resection for BCLM [3,26,27]. However, we found no evidence for such an association (p = 0.479), maybe due to the selected study cohort. In line with our results, Jakobs et al. [28] reported that hormone receptor status did not significantly affect survival after conventional RFA for BCLM. Late onset of BCLM has also been proposed as a predictor of survival by He et al. [29] and Hoffman et al. [30] in the surgical literature, although this is again distinct from findings in the RFA literature, including this study. One possible explanation for this difference could be due to the different selection criteria for hepatic resection vs. RFA. Another important reported prognostic factor of survival is the presence of extrahepatic disease at initial therapy, where we found extrahepatic metastases (excluding patients with isolated bone metastases) might be significantly associated with reduced survival (p = 0.025), which is an observation supported by Jakobs et al. [28] after conventional RFA.
Several studies [14,24,31] have also shown significantly reduced survival in patients with BLCMs [ 2.5 cm after RFA. Whilst larger tumors were associated with worse survival in our study, this did not reach statistical significance (p = 0.171, Table 4). In addition, Cox regression analysis revealed age \ 60 years was a positive predictor of survival, which is in line with the results of Dittmar et al. [32] following hepatic resection for BCLM.
Owing to the advantages of our setup, the technical success rate, i.e., accurate needle placement (deviation B 10 mm from the plan) was 100%. The measurement of the safety margin to determine success represents another important outcome measure, which we have not (yet) used in the present study. However, the importance of an adequate safety margin for local tumor control in patients with colorectal liver metastases and hepatocellular carcinoma has been addressed by our study group recently [33,34].
Insufficient local control remains a major drawback of conventional RFA, especially for larger tumors with reported local recurrence rates between 14 and 50% [14,35]. As such, to achieve complete ablation in larger tumors, multiple overlapping ablation volumes are required [36], although the resulting increase in complexity using  [37]. Besides RFA, microwave ablation (MWA) is an important ablative strategy which has a higher, and faster thermal energy transfer [38], which allows for larger ablation zones. However, studies regarding MWA for BCLM are sparse, with small patient numbers and ultimately inconclusive results. Whilst reported LR rates after conventional RFA for BCLM are between 11.6 and 25% [14,24,25], the majority of lesions are \ 3 cm. In comparison, our reported LR rate of 7.3% compares favorably to these results, given a median tumor size of 3 cm, which we attribute to our scrupulous technique, with the aim of achieving a sufficient ablation margin of at least 5 mm. This is achieved through the use of a sophisticated aiming device with precise needle placement and planning software which also allows fusion with previously acquired MR images in case of poor target visibility. Immediate post-ablation contrast-enhanced CT fusion with planning CT allows rapid and reliable assessment of ablation results with the option of repeat ablation in the same session. This standardized approach to ablation means tumors can be reliably treated without limitation in size [39] (the largest lesion treated in this cohort was 9 cm) and number [40]. Finally, in our experience, SRFA is easier to learn than conventional techniques, because important workflows can be trained under laboratory conditions, and the learning curve can be ascended before use in patients.
The mortality and complication rate (Grade 1 & 2) in our study were 0% and 10.4% (5/110), respectively, which is considerably higher than reported complication rates for conventional ablation, which range from of 0 to 1.1% [14,24,41]. The explanation for this is very likely to be due to the higher complexity of interventions, whereby tumors are often large and/or multiple. However, 4 out of 5 of the complications were successfully treated in the same anesthetic session and did not change the postoperative course. Nevertheless, our results compare favorably to hepatic resection for BCLM, which confers 0-5.9% mortality and 15% major morbidity [22].

Limitations
The limitations of this study include its retrospective design and a relatively small sample size. This small sample size reduces the accuracy of the subgroup analyses in particular. Use of additional therapies, such as chemotherapy and/or hormone therapy after SRFA should also impact the overall clinical outcome. Furthermore, comparison with previous studies is limited as stereotactic navigation systems were not employed in prior reports.
In conclusion, SRFA is a safe, minimally invasive treatment option in the management of BCLMs for selected patients who might benefit from local treatment, with similar survival rates to hepatic resection.
Author Contributions PS and BR contributed to study concepts, study design and data analysis. PS contributed to statistical analysis and manuscript preparation. DP, GL, GE, YS, CI, PS and BR contributed to data acquisition. DP, GL, GE, YS, CI, EW and BR contributed to manuscript review. EW and PS also involved in manuscript editing.
Funding Open access funding provided by University of Innsbruck and Medical University of Innsbruck. All authors have no conflicts to declare.

Declarations
Conflict of interest All authors have no conflicts of interest to disclose.
Consent for Publication Consent for publication was obtained for every individual person's data included in the study.
Ethical Approval All procedures performed in studies involving human participants were in accordance with the ethical standards of the institutional and/or national research committee and with the 1964 Helsinki declaration and its later amendments or comparable ethical standards.
Informed Consent Informed consent was obtained from all individual participants included in the study.
Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article's Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article's Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons. org/licenses/by/4.0/.