Hepatic resection for breast cancer related liver metastases: A single institution experience

Breast cancer represents a significant global health problem and is currently the most frequently diagnosed cancer and leading cause of cancer death in females. There was an estimated 2,088,849 new cases of breast cancer and 626,679 deaths from the disease worldwide in 2018.¹ A significant proportion of these patients will eventually develop distant metastases which is associated with an extremely poor prognosis. Approximately 50% of patients with metastatic breast cancer will develop hepatic metastases and the life expectancy for this cohort of patients is significantly reduced with some studies demonstrating a median survival of 2–21 months and 5–year survival rates ranging from 3.8% to 12%, although this varies depending on several patient-related factors and tumor biology.^2–4

The treatment paradigm for patients with metastatic breast cancer has undergone significant changes in recent decades. Cytotoxic chemotherapy has been complemented with or, in many cases, replaced by hormonal therapy, targeted therapy and more recently in selected cases, immunotherapy.^5–7 There is emerging evidence regarding the role of surgery and ablative therapies for women with metastatic breast cancer confined to a single site, although, it is clear that only a small fraction of women with metastatic disease are suitable for this approach.⁸ A number of small retrospective studies have reported results for resection of breast cancer-related liver metastases (BCLM) reporting median survival times of 30–70 months and 5-year overall survival rates of 33%–50%.^9,10 While resection of colorectal cancer liver metastases (CRCLM) is now the recommended gold standard for those with potentially resectable disease, there is still a lack of clarity and consensus on how best to treat those with potentially resectable BCLM.¹¹ In particular, there are no completed randomized controlled trials comparing surgical resection of BCLM to best available systemic therapy although one such trial has commenced recruitment.¹² There is however, a propensity score matching study demonstrating that surgery for BCLM yields a survival benefit compared to systemic therapy alone, particularly in patients with hormone receptor positive cancers.¹³ As many centers push ahead with surgical resection for BCLM we must ask ourselves which patients should be considered for this option and which patients should be spared the potential morbidity of liver resection and treated with systemic therapies instead.

The aim of this study was to describe the outcomes of the first series of patients in our institution to undergo hepatic resection for BCLM. Furthermore, we aimed to identify any clinicopathological variables that were associated with an increased risk of local recurrence in the liver or worse overall survival.

This study has described the outcomes for 20 patients with BCLM who underwent a liver resection in our institution over a 12-year period. The results demonstrate that liver resection is feasible and has an acceptable safety profile for selected patients with BCLM. Furthermore, this study has demonstrated that some patients can achieve excellent long-term results after liver resection for BCLM, with a median interval to hepatic recurrence of 50 months and an estimated 5-year OS of 65%. There is of course some selection bias as patients deemed suitable for liver resection were likely those with a pattern of disease and tumor biology that would have had an excellent prognosis with optimal medical management without surgery.⁸ One must also acknowledge the fact that only a minority of patients with metastatic breast cancer arrive at a point where they are deemed suitable for liver resection, demonstrated by the fact that only 20 patients were operated on over a 12-year period in our institution. This small number is consistent with several other previous studies.^14–16 It may also be true that many patients with BCLM are not referred to specialist hepatobiliary centers for consideration for liver resection due to a lack of acceptance of liver resection as part of the treatment paradigm for BCLM. Our initial experience with liver resection for BCLM has highlighted some important findings. First, we included a small cohort of patients with extrahepatic metastases (all bone metastases) that were deemed to be controlled, as several studies had not identified this to be an important prognostic factor.^{13,15,17–20} In our cohort all 4 patients with a history of extrahepatic metastases prior to liver resection were found to have local recurrence in their liver within 7 months of undergoing surgery and were also found to have a significantly worse overall survival. Given that our numbers were small, the outcomes for this cohort of patients requires further investigation to determine if liver resection can provide them with any meaningful benefit. Second, we could not identify any association between other clinicopathological factors with recurrence or survival. This means that based on our findings, patients with multiple metastases, bilobar metastases and large metastases should still be considered as potential candidates for surgical resection.

There are many factors beyond surgery which will influence outcomes for patients with BCLM. One of the most important factors is tumor biology which includes but is not limited to hormone receptor status and HER2 status. Hormone receptor positive and HER2 positive tumors provide oncologists with clear molecular targets with multiple therapeutic agents available for these targets.²¹ With currently available treatments, patient with relapsed or de-novo metastatic HER2 positive breast cancer can achieve median survival times in excess of 57 months⁵ and there is some evidence to suggest that HER2 positive breast cancer with liver metastases can be cured with a combination of chemotherapy and anti-HER2 agents.²² It is worth noting that the two patients who had a complete response in their liver to preoperative treatment were both HER2 positive. Those patients with TNBC still present a therapeutic challenge and are known to have a poor prognosis.²³ Our cohort had two patients with the TNBC subtype in their primary breast cancer, one of whom developed extrahepatic metastases after liver resection. A third patient was found to have TNBC in one of her two resected liver lesions despite the primary breast cancer being ER + ve/HER2-ve and she went on to develop local recurrence in her liver. Finally, a fourth patient with a HER2 + ve primary breast cancer and a HER2 + ve liver metastasis developed local recurrence in her liver after surgery requiring repeat resection and the recurrent lesion was found to be of the TNBC subtype. With such biological diversity the cohort of patients that are offered resection are extremely heterogeneous and this is shown not just by the biology of their tumors but also by the vastly different treatment regimens that each patient has completed by the time they make it to surgery. Several patients in our cohort have been treated with or are being considered for immunotherapy based on positive expression of programmed death-ligand 1 (PD-L1). Clearly comparisons of patients undergoing liver resection for BCLM is difficult when one considers the sheer number of confounding variables created by tumor biology and personalized approaches to treatment alone. Molecular subtype appears to influence outcomes after surgical resection, and patients with HER2-enriched subtypes have been shown to have prolonged overall survival after surgery compared with basal-like and luminal A subtypes.²⁴

Determining whether or not surgery for BCLM offers a survival benefit has proven difficult as there has never been a randomized controlled trial of surgery and optimal medical management versus optimal medical management alone. The BreCLIM-2 study comparing surgical intervention in the form of liver resection, ablation or stereotactic body radiotherapy with best oncological treatment to best oncological treatment alone has commenced recruiting patients for randomization, however the recruitment period is expected to run until December 2029 and with the primary outcome being survival three years after randomization, we may not see results from this study until 2033.¹² Furthermore, many studies focusing on surgical and medical management are confounded by the use of other techniques such as radiofrequency ablation, microwave ablation, stereotactic body radiotherapy and hepatic arterial infusion of chemotherapy.^25–28 A propensity score matching study has demonstrated a survival benefit for patients undergoing liver resection, particularly for those with hormone receptor positive disease.¹³ Two case matched studies have demonstrated improved survival for patients undergoing liver resection with systemic treatment versus systemic treatment alone, although another case control study has shown no difference in progression free survival between those undergoing surgery and those receiving systemic treatment only.^29–31 It seems plausible that a select cohort of patients will achieve a survival benefit from undergoing liver resection for BCLM and a recent meta-analysis has demonstrated a survival advantage for patients undergoing surgery combined with systemic therapy versus those receiving systemic therapy alone.³² Several other reviews have produced conflicting reports about the evidence of hepatic resection for BCLM.^33–35 There is also some evidence that this is a cost-effective treatment option when compared with systemic therapy alone, particularly in those with hormone receptor positive tumors and when newer systemic agents are used.³⁶ As the evidence for this therapeutic option increases we are beginning to see the boundaries being pushed further with new evidence suggesting that there may be a benefit in repeat liver resection for patients with recurrent liver disease after their first resection.³⁷ This raises questions as to whether or not some patients with BCLM will be treated in a similar fashion to patients with CRCLM, some of whom undergo two stage hepatectomies with portal vein embolization in an effort to achieve long-term survival.³⁸ There is also emerging evidence demonstrating that minimally invasive surgery is a safe and feasible option for patients undergoing hepatic resection for BCLM.³⁹

The main limitations of this study are the fact that it is retrospective in nature and that the cohort of patients included is small and therefore highly selected. The small number of patients makes it difficult to ascertain if any clinicopathological factors are truly associated with recurrence or survival, although we did identify an association between extrahepatic disease with earlier local recurrence and worse overall survival. We believe these findings are biologically plausible and these factors should be considered at the MDT meeting when deciding whether or not to offer surgery. There is clearly selection bias with a study of this nature and it is likely that we have offered liver resection to patients who would have had reasonably good long-term outcomes with optimal systemic therapy alone. The main strength of the study was that we were able to gather complete clinicopathological, intraoperative, treatment and follow up data on all included patients.

There have been significant gains made in the management of patients with metastatic breast cancer in recent years, particularly for those with hormone receptor positive and HER2 positive disease. There are several options available for patients with BCLM including liver resection. This study has shown that liver resection is safe and feasible in a select group of patients with BCLM and that long-term disease free and overall survival can be achieved in many of these patients. While BCLM is a different disease to CRCLM, we are beginning to see more widespread use of liver resection for BCLM in a similar fashion to CRCLM, including the use of repeat resection for those with local recurrence.