Temozolomide in secondary prevention of HER2-positive breast cancer brain metastases

Abstract

Brain metastases occur in up to 25–55% of patients with metastatic HER2-positive breast cancer. Standard treatment has high rates of recurrence or progression, limiting survival and quality of life in most patients. Temozolomide (TMZ) is known to penetrate the blood–brain barrier and is US FDA approved for treatment of glioblastoma. Our group has demonstrated that low doses of TMZ administered in a prophylactic, metronomic fashion can significantly prevent development of brain metastases in murine models of breast cancer. Based on these findings, we initiated a secondary-prevention clinical trial with oral TMZ given to HER2-positive breast cancer patients with brain metastases after recent local treatment in combination with T-DM1 for systemic control of disease. Primary end point is freedom from new brain metastases at 1 year. (NCT03190967).

Breast cancer is the most common cancer in women [1], and the second most common cause of brain metastases in USA [2]. The incidence of brain metastases varies among the different molecular subtypes of breast cancer, with the highest incidence in the HER2-positive and triple-negative breast cancers [3,4]. In the HER2-positive group, the diagnosis of brain metastases became even more frequent – up to 25–55% [5–9] – after widespread treatment with trastuzumab, a monoclonal antibody that improved control of systemic disease and survival but because it has low central nervous system penetrance did not prevent the development of brain metastases [10,11]. New agents targeting HER2 have been evaluated since the introduction of trastuzumab, and, though several are effective for systemic control of the disease, a major effect in decreasing brain metastases incidence in this group is still lacking.

Introduction to the trial

Current treatment of HER2-positive brain metastases

Radiation therapy modalities are currently the main options for treatment of brain metastases. In patients with limited number of brain metastases, current treatments include surgical resection, stereotactic radiosurgery (SRS) and/or whole-brain radiotherapy (WBRT) [12]. Most prospective trials evaluating local treatment of brain metastases included mainly NSCLC patients and only a small proportion of breast cancer patients [13–15]. Addition of WBRT to initial surgery or SRS has repeatedly demonstrated a decrease in intracranial disease recurrence but no difference in the overall survival of 7–10 months. However, WBRT has been reported to worsen quality of life and neurocognitive function, especially in patients with prolonged survival [16–18]. In those cases, neurocognitive decline is progressive and untreatable, though preventive strategies such as memantine and hippocampal avoidance have shown improvement this outcome [19,20]. Nevertheless, WBRT remains the main indicated treatment for patients with higher numbers of brain metastatic lesions, while adjuvant SRS is the standard of care following brain metastasectomy with limited number of lesions.

Systemic therapy for brain metastases, overall, has shown less efficacy than in systemic, non-CNS locations. Multiple clinical trials have documented few or no responses using agents with known activity in the systemic metastatic setting [21–26]. The reasons for this may be multifactorial, but likely includes the blood–brain barrier (BBB), which prevents most substances from exiting the bloodstream and entering the brain parenchyma or actively transports these agents back into the blood . Once a metastasis forms, the BBB is transformed into a blood–tumor barrier (BTB). In general, the BTB permits greater drug penetration than the BBB, but drug penetration (delivery) is heterogeneous and ∼1 log less than those reached in systemic metastases and therefore often ineffective [27].

In the HER2-positive breast cancer patients, HER2-targeted agents beyond trastuzumab have been also evaluated for their potential therapeutic effect in brain metastases. Lapatinib is a small molecule tyrosine-kinase inhibitor of EGFR and HER2, able to cross the BTB to a greater extent than paclitaxel [27,28]. A presurgical study documented heterogeneous levels of lapatinib in brain metastases from patients dosed pre-operatively for medically needed craniotomies [29]. As a single agent, lapatinib showed a few responses and a small nonsignificant diminution in size of brain metastatic lesions [30,31]. The combination of lapatinib plus capecitabine produced objective responses in approximately 30% of patients in Phase II trials evaluating the combination in patients previously treated with WBRT [31,32].

The LANDSCAPE trial evaluated the combination in a single-arm Phase II design as first-line treatment for low-volume brain metastases with objective CNS responses of 65.9% (measured by volumetric reduction), median time to CNS progression of 5.5 months and median time to WBRT of 8.5 months [33]. More recently, the MA.31 trial randomized 652 patients with HER2-positive breast cancer to treatment with lapatinib plus taxane, or trastuzumab plus taxane as first-line treatment of metastatic disease [34]. The primary end point was progression free survival (PFS). The trastuzumab combination was superior to the lapatinib combination, median 9.0 months and 11.3 months, respectively (HR 1.37; p = 0.001). The incidence of brain metastases as first site of progression was 28% for trastuzumab and 20% for lapatinib, with no difference in time to progression between the arms. In a brain metastasis prevention setting, a Phase III trial of lapatinib plus capecitabine versus capecitabine alone in patients with HER2-positive advanced breast cancer who were previously treated with an anthracycline, taxane, and trastuzumab was conducted. Four (2%) patients developed symptomatic brain metastasis as an initial site of progression in the combination therapy arm compared with 13 (6%) patients in the monotherapy group (p = 0.045) [35].

Pertuzumab is a monoclonal antibody that targets HER2 extracellular dimerization domain and acts in synergy with trastuzumab. It was demonstrated to prolong survival when used with trastuzumab and docetaxel as first-line treatment for metastatic HER2-positive breast cancer, compared with trastuzumab, docetaxel and placebo [36]. An exploratory analysis of the trial evaluated the incidence of brain metastases as first site of disease progression, and found it to be similar in the pertuzumab arm and the placebo arm (13.7 and 12.6%) [37].

Trastuzumab emtansine (T-DM1) is a newer anti-HER2 therapeutic agent, with case reports describing activity in CNS metastases [38,39]. T-DM1 is an antibody–drug conjugate containing emtansine (DM1), a microtubule-inhibitory agent, linked to trastuzumab. Phase III trials have shown activity of T-DM1 in HER2-positive metastatic disease after previous lines of treatment including trastuzumab and lapatinib, with an increase in PFS and OS [40,41]. T-DM1 is currently the standard therapy for HER2-positive breast cancer patients with recurrence or progression of disease after treatment with trastuzumab and pertuzumab. The EMILIA trial randomized advanced HER2-positive breast cancer patients previously treated with trastuzumab and taxanes, to receive either T-DM1, or the combination of lapatinib plus capecitabine [41]. Final results showed better PFS (median 9.6 months with T-DM1 vs 6.4 months with lapatinib plus capecitabine – hazard ratio for progression or death from any cause, 0.65; 95% confidence interval [CI], 0.55–0.77; p < 0.001), and OS (30.9 vs 25.1 months; hazard ratio for death from any cause, 0.68; 95% CI, 0.55–0.85; p < 0.001) for T-DM1. In a retrospective exploratory analysis of the EMILIA trial, the rate of CNS progression was similar between the two arms, with CNS metastases as the first site of relapse in 2% of T-DM1-treated patients and in 0.7% of lapatinib- plus capecitabine-treated patients, and progression of CNS disease known at baseline in 22.2 and 16%, respectively. However, in patients with treated asymptomatic CNS metastases at baseline, T-DM1 was associated with improved survival when compared with lapatinib plus capecitabine (median 26.8 vs 12.9 months; HR: 0.38; p = 0.008) [42].

Neratinib is an irreversible pan-HER tyrosine kinase inhibitor (HER1, HER2 and HER4) and received an orphan drug designation for HER2-positive breast cancer brain metastases by FDA in 2019, based on the Phase II TBCRC 022 trial [43]. The trial evaluated the combination of neratinib plus capecitabine in two cohorts, depending on the previous use of lapatinib. Volumetric brain metastasis response was used for evaluation. In the lapatinib-naive cohort (n = 37), 18 (49%) patients had PR and 7 (19%) patients had SD for ≥6 cycles (4.2 months), with median PFS 5.5 months and OS 13.3 months. When RANO-BM was applied to evaluate responses, 9 (24%) patients had a PR in that cohort. The cohort of patients that had received lapatinib in the past (n = 12) was closed for slow accrual, 4 (33%) patients had a PR and 3 (25%) patients had SD for ≥ 6 cycles (4.2 months), with PFS 3.1 months and OS 15.5 months. Per RANO-BM evaluation, 2 (17%) patients had a PR. Important toxicity was observed in both groups, with grade 2 and 3 toxicity levels incidence such as: 62% diarrhea (despite prophylaxis therapy), 24% nausea, 20% vomiting and 26% fatigue. Overall, neratinib showed some effect, mostly short lived, with non-negligible toxicity.

Tucatinib is a reversible tyrosine kinase inhibitor, that selectively inhibits HER2 and showed promising activity in combination with capecitabine and trastuzumab in Phase I trial, including responses in brain metastases [44]. Based on that, the HER2CLIMB Phase III trial was developed, with results recently reported [45]. This trial randomized 612 patients with HER2-positive metastatic breast cancer previously treated with trastuzumab, pertuzumab and T-DM1, to trastuzumab and capecitabine, combined with tucatinib versus placebo. Innovatively, patients with brain metastases were allowed either without or after local therapy when indicated for symptom control. The addition of tucatinib improved PFS at 1 year (33.1 vs 12.3%, HR 0.54; 95% confidence interval [CI], 0.42 to 0.71; p < 0.001), with the median duration of PFS 7.8 and 5.6 months. The improvement also reflected significantly on OS at 2 years (44.9 vs 26.6%; HR: 0.66; 95% CI: 0.50–0.88; p = 0.005), with median overall survival of 21.9 and 17.4 months. Patients with brain metastases had 1-year PFS of 24.9% with addition of tucatinib and 0% in the placebo-combination group (HR: 0.48; 95%: CI; 0.34–0.69; p < 0.001), with median PFS 7.6 and 5.4 months, respectively. This represents an important effect, however, details on the brain metastases specific response and PFS in brain are still to be reported.

Independent of the localized or systemic treatment modality, once brain metastases are established, treatment options are currently limited and the disease almost invariably progresses, limiting not only survival but also quality of life in most patients.

Background & rationale

Prevention of brain metastases

Preventing the development of brain metastases in high-risk breast cancer patients is a potentially novel approach. Primary prevention refers to the initial development of a brain metastasis; secondary prevention refers to the development of additional brain metastases in patients with limited metastases treated with local therapy.

The preclinical literature suggests the hypothesis that preventing the formation of a metastasis by a drug may be more efficacious than attempting to shrink an established lesion. In this context, multiple mouse models of brain metastasis of breast cancer have been developed enabling the testing of many approved drugs and candidates in a prevention mode, where they were delivered early and continuously after tumor cell injection in a hematogenous model. Alternatively, drugs have been tested in a therapeutic setting, where brain lesions were permitted to form, either hematogenously or by intracranial implant, and then treatment began. As reviewed in [46], several drugs have shown statistically significant efficacy in preventing the formation of brain metastases. Far fewer drugs have shown treatment activity, generally limited to models involving intracranial implantation of tumor. This low level of therapeutic efficacy is in keeping with the clinical literature.

The use of prophylactic radiation therapy, as used in small-cell lung cancer, is one approach to prevent brain metastases in high-risk patients. At least two trials, one in Europe (NCT00639366) and other in Canada (NCT00916877), planned to recruit high-risk breast cancer patients to evaluate prophylactic radiation and prevention of brain metastases. NCT00639366 reported negative results after accruing only 51 patients of the initially planned 390 patients [47]. However, the increasing evidence of neurocognitive decline related to WBRT and the lack of initial reports showing major sign of positive results are discouraging [47]. There is increasing evidence of at least transient declines in memory, executive function, and processing speed in the first few months after WBRT in up to 40–60% patients. Based on previously published results, additional progressive neurocognitive complications may be expected to develop with time.

In an attempt to evaluate prophylactic treatment against brain metastases as first site of relapse, the Phase III CEREBEL trial assessed 501 women that were randomized to lapatinib plus capecitabine versus trastuzumab plus capecitabine [48]. However, the incidence of brain metastases was overall low, 3 and 5% respectively, with no significant difference between the two arms.

Considering the different subtypes of breast cancer, a retrospective analysis of 154 patients (56% HER2-positive) treated with SRS, evaluated their outcome in terms of new brain metastases, overall survival and death ultimately caused by progression in brain [49]. Median overall survival was 16.7 months for HER2-positive patients and 8.4 months for HER2-negative patients (p < 0.001). Brain metastasis progression was the cause of death in 46% of HER2-positive and 31% of HER2-negative patients (p = 0.066). Freedom from distant brain failure at 1 year was 58% in HER2-positive as opposed to 40% in HER2-negative patients (p = 0.034), however, the majority of the patients in both groups had also received WBRT in the past. This analysis reinforces the findings that, once brain metastases are found, HER2-positive patients may have a better overall survival, but they are more prone to develop and to die from brain metastasis progression. Most interestingly, since brain metastatic disease frequently recurs after initial SRS, a secondary prevention strategy could be helpful at this point.

Innovative preclinical assessment

Using models developed with brain tropic subclones of the HER2–positive JIMT-1-BR3 (in vitro) and the triple-negative 231-BR-EGFP (in vitro and in vivo in animal models) cell lines, our group has shown that, even in very low doses, Temozolomide (TMZ) administered in a prophylactic fashion can prevent development of brain metastases (Table 1) [50]. The brain tropic tumor cells were injected into left cardiac ventricle and on day 3 the mice were randomized to vehicle versus TMZ given by oral gavage for 5 days every week. At necropsy, brain metastases were quantified in step-wise sections through one hemisphere, and doses of TMZ from 5 to 50 mg/kg significantly prevented brain metastases formation. TMZ was ineffective in the treatment (shrinking) of established brain metastases. When mice injected with brain-tropic breast cancer cells were treated with a course of TMZ and then allowed to live untreated, TMZ extended survival, with early treatment better than later treatment. TMZ prevention of brain metastases was dependent on relatively low tumor expression of O6-methylguanine-DNA methyltransferase (MGMT), an enzyme that is the main mechanism of resistance to TMZ. In a comparison of matched primary breast tumors and resected brain metastases, HER2-positive disease showed the least MGMT expression in brain metastases. The ability of TMZ to prolong life more effectively when given earlier is consistent with its ability to prevent metastases but inability to treat established metastases.

Table 1. . Prevention of experimental brain metastases by temozolomide over a wide dose range.

Experiment	Dose (mg/kg)	Median number metastases per brain section
		Large	Micro
1	0	2.0	63.3
	50	0	0
2	0	2.6	86.4
	50	0	0
	25	0	0
3	0	6.5	143.3
	25	0	0
	10	0	0
	5	0	0
4	0	2.3	101.1
	1	0.9	40.1
	0.5	1.2	58.0

Mice were injected with brain tropic MDAMB231-BR cells and treated from day 3 post-injection with the indicated dose of TMZ. At necropsy large (>300 um) and micrometastases were quantified in step sections through one brain hemisphere. Experiments 1–3; p < 0.01. Results of four experiments are shown to demonstrate the efficacy of TMZ dose.

Adapted with permission from [50].

Changing paradigm

We proposed a secondary prevention clinical trial with oral TMZ for HER2-positive breast cancer patients with 1–10 brain metastases after recent local treatment (SRS or surgical resection), with the primary end point to decrease incidence of new metastatic lesions in the brain at 1 year. TMZ is combined to the anti-HER2 agent T-DM1 for systemic control of disease.

Design

Study design & outcome measures/end points

The proposed study is a Phase I/II open label study to evaluate the potential benefit of TMZ in prevention of new brain metastases in patients with limited brain metastases from HER2-positive breast cancer, previously treated with SRS or surgical resection of brain metastases. Figure 1 (study schema). All patients receive the standard second-line therapy for HER2-positive metastatic breast cancer: T-DM1. During Phase I, patients will receive both T-DM1 in standard dose and TMZ in increasing dose levels. During Phase II patients will be randomized between T-DM1 plus TMZ versus T-DM1 alone, at the maximum tolerated dose (MTD) determined in the Phase I.

• Phase I run in: T-DM1 3.6 mg/kg IV every 21 days plus TMZ 30, 40 or 50 mg/m² daily;

• Phase II: T-DM1 3.6 mg/kg versus T-DM1 3.6 mg/kg plus TMZ at recommended Phase II dose (RP2D).

Figure 1. . Study schema.

SRS: Stereotactic radiosurgery; T-DM1: aAo-trastuzumab emtansine.

The primary end point of Phase I (run in) is to establish the maximum tolerated dose (MTD) of TMZ when used in combination with T-DM1. The primary end point of Phase II is to determine if the combination regimen of T-DM1 and TMZ improves the freedom from distant new brain metastases following stereotactic radiosurgery or surgical resection in HER2-positive breast cancer brain metastases at 1 year as compared with T-DM1 alone.

Drug combination

Temozolomide

To our knowledge, the only published trial evaluating the combined use of TMZ with an anti-HER2 agent in breast cancer brain metastases was the LAPTEM trial [51], combining TMZ with lapatinib for treatment of established brain metastases, not prevention. The regimen showed a favorable toxicity profile, with main side effects being fatigue, diarrhea and constipation. TMZ was used in three dose levels: 100, 150 and 200 mg/m²/day, days 1–5 with cycles of 28 days; lapatinib was given orally, once a day also at three dose levels: 1000, 1250 and 1500 mg/day. The MTD was not reached, per protocol. No cardiac toxicity events were observed. Out of the 15 evaluable patients, 10 achieved stabilization of disease, with estimated median survival time of 10.94 months (95% CI: 1.09–20.79), and median progression-free survival time of 2.60 months (95% CI: 1.82–3.37). Pneumocystis jirovecii pneumonia prophylaxis is recommended to patients in use of TMZ with CD4 counts <200/mm³, based on persistent lymphopenia associated to some of the glioblastoma regimens [52].

Trastuzumab-emtansine (T-DM1)

T-DM1 is currently the standard second-line therapy for treatment of metastatic HER2-positive breast cancer, after the use of trastuzumab plus pertuzumab, as discussed in detail above. T-DM1 has also become standard therapy in the adjuvant setting for patients with HER2-positive breast cancer with residual tumor after neoadjuvant chemotherapy, based on KATHERINE trial evidence [53].

Eligibility criteria

Main eligibility criteria are summarized on Table 2.

Table 2. . Selected eligibility criteria.

General Histologically confirmed HER2+ breast cancer Brain metastases recently treated (local therapy) within 12 weeks of study entry ECOG performance status 0–2 and adequate organ and marrow function Patients with systemic metastases are eligible Patients with leptomeningeal metastatic disease are ineligible Patients that are unable to complete a brain MRI with contrast are ineligible Patients who have received previous treatment with T-DM1 and had systemic progression of disease while on it, are ineligible. Patients receiving treatment with T-DM1 whose only site of disease progression was brain are allowed to enroll in this trial

PHASE I specific Any number of brain metastases is allowed. WBRT is allowed

PHASE II specific Up to 10 brain metastases, by contrast MRI, treated within 12 weeks of study entry with SRS and/or resection Patients with history of WBRT are ineligible

Planed sample size & study period (Statistics)

Phase I will follow a standard 3+3 design. Thus, with three dose levels, up to 18 patients may be included in the initial safety evaluation. The patients for the Phase I will be enrolled and treated at the NIH in Bethesda, MD. In the Phase II portion of the trial, a total of 49 evaluable subjects per arm (98 total) will need to be randomized over a 3-year period and followed for an additional 2 years from the date of entry of the last patient, with occurrence of 79 total relapses in the brain in both arms combined, in order to have 80% power to compare the curves with a 0.10 significance level one-sided log rank test. Stratification per number of brain metastases, presence or not of systemic metastases, and previous adjuvant treatment with T-DM1 is planned. The Phase II should be available to patients in multiple centers.

Study procedures

Clinical evaluation and laboratory tests (including CBC, biochemical profile and CD4 count) will be performed to evaluate for eligibility and follow-up every 3 weeks during treatment. Baseline disease characteristics, previous treatment, medical history and demographics will be collected at screening visit, as well as physical examination data. Metastatic breast cancer assessment will be done by chest, abdomen and pelvis CR scan and brain MRI every 6 weeks (±7 days) while on treatment in the trial. Tumor measurements will be evaluated locally and centrally per RECIST for systemic disease and RANO-BM criteria for CNS disease. Echocardiogram or MUGA-scan will be performed every 3 months during use of T-DM1. Lumbar puncture will be performed by Interventional Radiology on C1D1 of treatment and on C3D1 for collection of CSF samples. Dose delays, reductions and discontinuations due to adverse events (AEs) are defined in the protocol. Patients receiving anti-hormonal therapy will be allowed to continue it. An exploratory subgroup analysis is also planned for this group. Clinical outcomes in this study will be evaluated using patient reported outcomes (PROs) including the MDASI-BT [54] and PROMIS^® cognitive function will be utilized for evaluation of overall symptom burden and perceived cognitive symptom and function Both questionnaires will be filled by the patients at every imaging, before the appointment with the physician.

Correlative studies

In an attempt to expand our knowledge on breast cancer brain metastases, studies will address aspects of TMZ resistance, molecular alterations in brain metastatic tumor cells, the neuro-inflammatory response and markers of cognition.

DNA sequencing

Few molecular tools exist to analyze the evolution and impact of brain metastasis evolution other than MRI scans and patient outcomes testing; tumor tissue evaluation is anatomically limited. Detecting DNA alterations in brain metastasis from patients’ systemic samples will likely be compromised by the contributions of systemic disease. Recently, several publications have reported the sequencing of cell free DNA (cfDNA) in CSF from patients with brain metastases [55]. The presence or absence of cfDNA in CSF, or its levels and alterations was a correlate of clinical course and more informative than serum cfDNA [56]. Moreover, recent work from the Brastianos laboratory, demonstrated that brain metastases show clinically actionable mutations not detected in the clinically sampled primary tumors [57]. Therefore, we propose sequencing of cell free tumor DNA in the CSF as an exploratory end point in our clinical trial. And we hypothesize that: patients will have mutations that can be identified in the CSF in the secondary brain metastasis prevention trial; alterations in the pattern or prevalence of mutations will be associated with clinical outcome in the secondary brain metastasis prevention trial and that alterations in the pattern or prevalence of mutations or cfDNA levels in CSF will predict relapse before MRI imaging. In addition, the data will provide a landscape of the genetic mutations from cfDNA in CSF for brain metastasis of breast cancer, and a portrait of its evolution through progression. The CSF cfDNA sequence data will be compared with that of the primary tumor where available and serum cfDNA.

MGMT status

MGMT is a DNA repair enzyme that confers resistance to TMZ. MGMT expression and activity have been determined by several methods in clinical tissues, most often by the DNA methylation status of its promoter, including methylation specific PCR (MSP), pyrosequencing and semi-quantitative MSP. Other techniques quantify total MGMT expression including IHC and QRT-PCR [58].

In preliminary studies in the laboratory of Dr. Renata Duchnowska, MGMT expression was determined by IHC in FFPE blocks of matched sets of primary breast tumors and resected brain metastases (n = 62) [50]. Samples were dichotomized into low expression (<5% of tumor cells+) and high expression (>5%) categories by a pathologist. Overall concordance of primary tumor and brain metastasis MGMT expression was weak. Low MGMT expression was prevalent, in 58% of brain metastasis specimens. A strong trend was observed in brain metastases for an association of HER2 overexpression and MGMT negativity (p = 0.089), suggesting the eligibility of this subset for potential clinical trials.

The hypothesis to be tested in this trial is that MGMT status, either at baseline or changes over time, will reflect secondary preventive activity of TMZ. A second hypothesis posits that alterations in MGMT may be an early sign of resistance to therapy and precede progression detectable on MRI. In addition, we will determine the concordance of MGMT methylation from multiple sources (tissue, serum, CSF) and by two assays (methylation and IHC). MGMT status will be assayed from patient serum at baseline and on study, the primary FFPE tumor block as available, and CSF as available in this trial. MGMT methylation will be used for all samples. For the FFPE block, MGMT IHC will also be conducted as previously reported.

Neuroinflammatory response & neuronal damage metrics

Brain metastases elicit a prominent neuro-inflammatory response including activated astrocytes, identified by GFAP positivity and microglia [59]. Multiple preclinical studies indicate that the neuro-inflammatory milieu conducts cross-talk with tumor cells to augment metastatic colonization [60]. The hypothesis to be tested is that the extent of neuro-inflammatory response, either at baseline or through treatment, will correlate with secondary prevention efficacy of TMZ. Other goals of these studies are to determine the prevalence, magnitude and kinetics of the neuro-inflammatory response in brain metastasis progression, which has not been addressed in the clinical literature.

The neurocognitive sequelae of brain metastases derive from both the lesions and the treatments [61–63]. In preclinical studies, brain metastases of breast cancer caused neuronal cell death [64], which could contribute to neurocognitive dysfunction. Regional or systemic tests for neuro-inflammatory responses and neurocognitive sequelae have not been meaningfully incorporated into clinical research on brain metastasis, other than mental function testing. Several leads have been developed for other brain pathologies and will be tested in this trial in an attempt to validate markers such as GFAP and UCH-L1, developed in the traumatic brain injury field, as markers of neuro-inflammation and neuronal injury, respectively [65–67]. Quantification of GFAP and UCH-L1 will be performed in serum and CSF samples from all patients on the trial. MicroRNA-34c (miR-34c) has been studied in Alzheimer's disease (AD) as a biological covariate of cognitive function and will be evaluated in our trial as well [68].

Exosomes

Exosomes are small membrane vesicles (30–100 nm in diameter) of endocytotic origin and secreted by most cells constitutively. These microvesicles contain cell type-specific proteins and genetic materials, including mRNAs, miRNAs and DNA, and can exert a functional influence once taken up by recipient cells, therefore representing novel mediators of intercellular communication. Importantly, exosomes are shed from cancers and are readily isolated from the peripheral circulation. A growing body of evidence has shown that tumor-derived exosomes not only have a crucial role in regulating tumor growth and metastasis, but also provide a ‘treasure box’ of novel noninvasive biomarkers for different types of cancers [69]. The laboratory of Dr David Lyden recently reported that the exosomal CEMIP protein functionally promotes brain metastasis in preclinical models [70]. We propose to use this ‘liquid biopsy’ in this trial to evaluate the possibility of rapidly screening for metastatic disease detection, as well as for metastasis prediction and treatment response monitoring.

Conclusion & future perspective

Brain metastases continues to have a high incidence in patients with HER2-positive breast cancer, with a large impact in the lives of these patients. Despite the continuous development of new and advanced therapies and even some progress, this remains an unmet need in the field. Based on preclinical evidence, we described here the concept of a new clinical trial with metronomic use of TMZ for prevention of new brain metastases as a proof of concept that may potentially change this paradigm. The Phase I component of this trial, evaluating the safety of the combination of TMZ and T-DM1 in patients with HER2-positive brain metastases after local treatment is about to complete accrual (NCT03190967). The Phase II component of the trial will open for accrual once the Phase I is completed and it should be considered for secondary prevention of brain metastases.

Executive summary.

Breast cancer brain metastases

• Brain metastases occur in up to 25–55% of patients with metastatic HER2-positive breast cancer. Standard treatment has high rates of recurrence or progression, limiting survival and quality of life in most patients.

• Standard treatment with surgery and radiation has high rates of recurrence or progression, limiting survival and quality of life in most patients.

• Systemic therapy for brain metastases, overall, has shown less efficacy than in systemic, non-CNS locations.

• Multiple clinical trials have documented few or no responses using agents with known activity in the systemic metastatic setting [21–26].

• Independent of the localized or systemic treatment modality, once brain metastases are established, treatment options are currently limited and the disease almost invariably progresses, limiting not only survival but also quality of life in most patients.

Temozolomide in innovative preclinical assessment

• Temozolomide (TMZ) is cytotoxic alkylating agent with 100% oral bioavailability that penetrates the blood–brain barrier and is approved for glioblastoma multiforme and recurrent anaplastic astrocytoma. Mixed responses in brain metastases treatment several different tumors.

• A mouse xenograft model of breast cancer brain metastases demonstrated that TMZ administered in a prophylactic fashion can prevent development of brain metastases.

• TMZ was ineffective in the treatment (shrinking) of established brain metastases.

Changing paradigm: secondary prevention of brain metastases clinical trial

• We developed a secondary-prevention clinical trial with oral TMZ for HER2-positive breast cancer patients with brain metastases after recent local treatment (SRS or surgical resection)

• Primary end point is improvement in the rate of new metastatic lesions in the brain at 1 year.

• TMZ is combined to the anti-HER2 agent T-DM1 for systemic control of disease.

• A Phase I portion is currently open to evaluate safety of the concomitant use of TDM-1 and TMZ. Patients will receive both T-DM1 in standard dose and TMZ in increasing dose levels.

• During Phase II, patients will be randomized between T-DM1 plus TMZ versus T-DM1 alone, at the maximum tolerated dose (MTD) determined in the Phase I.

• Eligibility includes patients with histologically confirmed HER2-positive breast cancer, adequate liver and bone marrow function, with brain metastases treated with SRS and/or surgery until up to 12 weeks before trial enrollment, being eligible for treatment with T-DM1 in metastatic setting and with good performance status (ECOG 0-2). During Phase I patients with any number of brain metastases and treated with WBRT are allowed.

• During Phase I, the patients will receive T-DM1 at a standard dose of 3.6 mg/kg IV every 21 days plus TMZ 30, 40 or 50 mg/m² daily, following a standard 3+3 design. During Phase II, patients will be randomized 1:1 between T-DM1 3.6 mg/kg alone and T-DM1 3.6 mg/kg plus TMZ at recommended Phase II dose (RP2D).

• In the Phase II portion of the trial, a total of 49 evaluable subjects per arm (98 total) will be randomized and followed for an additional 2 years from the date of entry of the last patient, with occurrence of 79 total relapses in the brain in both arms combined, in order to have 80% power to compare the curves with a 0.10 significance level one-sided log rank test.

• Correlative studies will address aspects of TMZ resistance, molecular alterations in brain metastatic tumor cells, the neuro-inflammatory response, and markers of cognition.

• For more information, please visit: https://clinicaltrials.gov and search NCT03190967.