Data from MCC supports a significant rise in the number of BC-LMD cases between 2011–2020. Greater institutional efforts to identify and treat LMD patients were made in the latter half of the study period, which may have driven the increased in BC-LMD diagnosis. A slight trend towards increased proportions of patients with TNBC were seen in the current review, but all cases increased over the study period. One of the first to report this, showed that HR status influences the risk of developing LMD.12 This is supported by recent data suggesting HR + BC-LMD likely comprises the majority of all cases (range 48–66%), and rates of HER2 + and TNBC-LMD are more variable.12–16 HER2 + BC-LMD represents 14-47.4% of all cases, and TNBC-LMD varies between 13.1–40% across studies.12–16 The proportion of HR + patients was similar in our study at 51% compared to HER2 + BC-LMD (27%) and TNBC (22%).
BC subtype affected time between CNS metastasis and BC-LMD diagnoses. TNBC had the fastest progression to BC-LMD compared to HR + and HER2 + BC-LMD. Even though HER2 + cancer may have a predisposition for CNS invasion, a tendency to invade the leptomeninges has not been clinically revealed.1,17 It has been shown that increased survival in patients with BC generally correlates with increased incidence of CNS metastases.18 Prolonged survival and risk of developing metastases, may confer a greater risk of developing BC-LMD.14 Historically, HER2 + and TNBC have carried an increased risk of developing CNS metastasis, and an associated decreased survival relative to HR + disease.19,20 Earlier work suggested no difference among BC-LMD based on the molecular subtype.21 However, our findings support more recent studies demonstrating an increased risk of developing LMD for TNBC.12,13,15,16,22
This study also showed the use of any systemic treatment and/or WBRT post-CNS metastasis delays progression from to BC-LMD. HR + patients receiving hormone therapies and/or kinase inhibitors may experience prolonged times between CNS metastasis and LMD diagnosis. For HER2 + patients, treatment with lapatinib demonstrated similar results. However, the studies stringent criterion for stratifying patients by at least a one-month period between CNS metastasis and LMD severely reduced the total patient sample from 128 to 48 patients. Follow-up studies analyzing significantly larger patient samples is required to definitively determine factors that prolong time between CNS metastasis and BC-LMD.
Survival did not change over the study period. Five cases identified in the first two years of the study survived longer than two years, but survival over the decade remained static even accounting for these outliers. Median survival after the diagnosis of BC-LMD was 4.7 months in this study. TNBC LMD had the shortest median survival of 2 months, followed by 5.3 months in HR + and 8.4 months in HER2+. These findings concur with previous studiesand BC-LMD survival depending on BC subtype.22,23 While longer survival for patients with HER2 + BC-LMD likely stems from HER2 targeted systemic and IT chemotherapy.1,17,22,24 differences in survival in HR + or TNBC may be driven by other factors. Patients with HR + BC-LMD were five times as likely to receive systemic therapy and four times as likely to receive IT chemotherapy following BC-LMD diagnosis compared to those with TNBC. It is tempting to say, therefore, that intention to treat is the primary driver of prolonged survival, but other factors including poor performance status and extent of extracranial metastatic disease may be confounders in this data.23 Even so, we were unable to identify any other single factor accounting for differences in survival among patients with HR + or TNBC.
The use of systemic therapy and IT chemotherapy appeared to improve survival following the diagnosis of LMD regardless of HR subtype. IT trastuzumab showed prolonged survival in HER2 + LMD.22,25,26 In 2018, the first phase 1 study of IT trastuzumab showed 150 mg weekly dosing achieved steady-state levels after 1 week and was well tolerated.27 Prior to this publicationy, patients with HER2 + LMD treated at our institution received weekly IT doses of trastuzumab of less than 150 mg. Our study demonstrates that IT trastuzumab dosing (i.e. physiologic dosing) resulted in improved survival for patients with HER2 + disease.
To maximize IT HER2 targeted therapy, our institution opened a phase I/II study of radiotherapy followed by IT trastuzumab and pertuzumab in patients with HER2 + BC-LMD to evaluate safety and treatment outcomes (NCT04588545).24 Radiotherapy can eliminate tumor blockages within the leptomeninges, allowing for IT therapy to properly flow along the CSF.28 The current study, however, found no statistical difference in the OS between patients who received only IT therapy versus WBRT and IT therapy. However, the latter cohort included only four patients. Further investigation is warranted.
There is an increasing interest in HER2-targeting therapies in BC-LMD. Prior studies showed that pertuzumab-based therapies improved progression-free survival (PFS) when used with trastuzumab and taxanes, and when given as a first or second-line chemotherapy 29, 30y 31.
Of particular interest in the current study, ten HER2 + BC-LMD patients exhibited a significant increase in OS in association with systemic treatment with lapatinib, a reversible tyrosine kinase inhibitor (TKI).32,33 Lapatinib with capecitabine is well tolerated34 and improved PFS to 8.4 months versus 4.1 months using capecitabine alone.35 Lapatinib + capecitabine BC brain metastases showed overall response rates (ORR) of 59.1% among treatment naïve patients36 and 21% among patients who may or may not have been exposed to either agent previously.37 However, the CEREBEL trial demonstrated no difference in PFS between capecitabine plus either lapatinib or trastuzumab38.The EMILIA trial showed that lapatinib plus capecitabine was both less tolerable and less efficacious than trastuzumab emtansine in prolonging PFS.39 The LANTERN trial, a phase II trial comparing lapatinib-capecitabine versus trastuzumab-capecitabine therapy in HER2 + BC with CNS metastasis showed no significant difference in PFS but a trend favoring trastuzumab-capecitabine.40 Aside from two case studies, there is no prospective data describing the efficacy of lapatinib in the treatment of BC-LMD.41,42
Neratinib and pyrotinib are similar to lapatinib except that they irreversibly bind to the HER intracellular phosphorylase domain and have efficacy in BC brain metastases.33 The NALA trial compared lapatinib and neratinib and both with capecitabine in BC brain metastases and showed similar ORR (26.7% versus 32.8%, respectively), but significantly longer duration of response for neratinib rather than lapatinib (8.5 versus 5.6 months, respectively). The benefits of these TKIs were overshadowed by the combination of T-DM1 and tucatinib, as tolerability was better and efficacy was at least as good for these agents.16, 33 Neratinib + capecitabine enhanced OS to 10 months and improved neurological symptoms in 60% of patients with HER2 + BC LMD.43
Tucatinib is a newer TKIthat has shown activity in HER2 + BC brain metastases, when combined with trastuzumab and capecitabine.44–46 CSF pharmacokinetic analysis revealed detectable levels of tucatinib within 2 hours of administration (NCT03501979).47 Further studies using TKIs in BC-LMD are needed. Other studies evaluated these agents in CNS metastases and the ORR in BC-LMD was not established.
At the present time, there is a lack of effective treatments in TNBC-LMD. The current study showed an association between IT therapy and OS among BC-LMD patients, supporting prior studies.22 Similarly, we found that patients with TNBC-LMD were less likely to receive treatment after their LMD diagnosis. Prospective trials specifically targeting TNBC-related LMD are critically needed.
A general observation based on the current data conforms with previous work showing that treatment of any type following diagnosis of BC-LMD improved survival to 6.54 months.3, 22, 23 It might be surmised that greater intention to treat would improve survival, as this study found that patients that did not receive any treatment survived a median of 1.07 months. Survival has also been shown to vary when considering diagnostic modality: cytology versus MRI alone.2 However, we found no difference in OS based on the diagnostic modality. This contradiction may relate to our exclusion of cases of BC-LMD, which were treated on clinical suspicion of LMD but for which no CSF cytology or MRI evidence of disease was found.17 There may also be improving awareness and recognition of typical radiographic LMD features, which increases the sensitivity of MRI in our institution.
To expand systemic treatment affecting CNS and LMD-related cancer, various novel immunotherapy approaches are being assessed in the management of BC-LMD. The use of systemic pembrolizumab48 or systemic ipilimumab and nivolumab49 showed promise for a variety of LMD patients (most of whom had LMD from BC), but the median survival was only 3.6 months and 2.9 months, respectively. Only 11 patients at our institution received systemic immunotherapy following diagnosis with BC-LMD, and no associated benefit was found. An alternative approach in trials now includes IT bispecific antibody-armed T-cells that can be directed against HER2 (NCT03661424). A similar trial with IT HER2-directed chimeric antigen receptor (CAR) T cells (NCT03696030) is recruiting currently. A phase 3 study is planned to use the systemic administration of blood-CSF penetrant drug ANG1005 (a paclitaxel-peptide conjugate that crosses the blood CSF barrier via a low-density lipoprotein receptor-related protein-1 (LRP-1) mediated transcytosis). A phase 2 study with the same agent showed activity and an average survival of 8 months in BC patients with LMD.50 At the time of this report, there are only 23 recruiting or active clinical trials in the US targeting LMD generally and even fewer targeting BC-LMD. Even with improvements in outcomes for HER2 + BC-LMD, the need to find new therapies which improve OS in BC-LMD is dire.