Sage Journals: Discover world-class research

Abstract

Background:

The survival status of patients with breast cancer and brain metastasis (BCBM) receiving current treatments is poor.

Method:

We designed a real-world study to investigate using patients’ clinical and genetic aberrations to forecast the prognoses of BCBM patients. We recruited 146 BCBM patients and analyzed their clinical features to evaluate the overall survival (OS). For genetic testing, 30 BCBM and 165 non-brain-metastatic (BM) metastatic breast cancer (MBC) patients from Hunan Cancer Hospital, and 86 BCBM and 1416 non-BM MBC patients from the Geneplus database who received circulating tumor DNA testing, were compared and analyzed.

Results:

Ki67 >14% and >3 metastatic brain tumors were significant risk factors associated with poor OS, while chemotherapy and brain radiotherapy were beneficial factors for better OS. Compared with non-BM MBC patients, BCBM patients had more fibroblast growth factor receptor (FGFR) aberrations. The combination of FGFR, TP53 and FLT1 aberrations plus immunohistochemistry HER2-positive were associated with an increased risk of brain metastasis (AUC = 77.13%). FGFR aberration alone was not only a predictive factor (AUC = 67.90%), but also a significant risk factor for poor progression-free survival (Logrank p = 0.029). FGFR1 aberration was more frequent than other FGFR family genes in BCBM patients, and FGFR1 aberration was significantly higher in BCBM patients than non-BM MBC patients. Most FGFR1-amplified MBC patients progressed within 3 months of the late-line (>2 lines) treatment.

Conclusion:

A group of genetic events, including FGFR, TP53 and FLT1 genetic aberrations, and HER2-positivity, forecasted the occurrence of BM in breast cancers. FGFR genetic aberration alone predicted poor prognosis.

Keywords

breast cancer with brain metastasis circulating tumor DNA HER2-positive PFS and OS

Background

Breast cancer (BC) is the most common malignancy in females. Metastases to the brain occurs in 10–16% patients with BC.^1,2 Compared with hormone receptor (HR)-positive BCs, which are more likely to recur in bone, triple-negative BCs (TNBC) and HER-2 positive BCs more commonly recur in the brain.³ BC with brain metastasis (BCBM) is a devastating cause of morbidity and mortality. The mean interval time from primary BC diagnosis to BCBM existence is about 35 months. The two main risk factors for developing BCBM are large primary tumor size and lymph node metastasis.⁴ Clinically, brain metastases are treated with surgery, radiation therapy [whole brain radiation therapy (WBRT)] or stereotactic radiosurgery. In practice, no uniformly standard chemotherapy drugs are available for BCBM; in the National Comprehensive Cancer Center guideline, the category 2A recommendations include high-dose methotrexate, capecitabine, temozolomide monotherapy and cisplatin plus etoposide.⁵ Despite a lack of consensus, preliminary data suggests that chemotherapy and targeting therapies after WBRT may improve survival outcomes.⁶

The majority of BCBM patients receive a multimodality therapy approach, including anti-HER2 therapy in HER2-positive tumors and hormonal therapy in estrogen receptor (ER) and/or progesterone receptor (PR) positive tumors.⁷ Trastuzumab and lapatinib, administered concurrently or sequentially, may improve overall survival (OS) in HER2-positive BCBM patients.^8,9 Compared with an OS of 16.7 months for a trastuzumab-based regimen alone, trastuzumab plus lapatinib and capecitabine show a significantly longer OS of 27.9 months (p = 0.01).¹⁰ The targeting therapy options for triple-negative BCBM are currently not available in clinical practice. For some triple-negative BCBM patients who harbor BRCA1 and BRCA2 aberrations, PARP inhibitor olaparib has demonstrated activity in BRCA1/2(MUT+) BC.¹¹ In addition, quite a few advanced clinical trials for new targeting agents are in process, including EGFR/VEGFR inhibitors, PARP inhibitors, PIK3CA/mTOR inhibitors and CDK4/6 pathway inhibitors.^12–15 Targeted therapy with anti-PD-L1 antibody atezolizumab in PD-L1 positive-TNBC is a new standard of care.¹⁶ Compared with placebo plus nab-paclitaxel, atezolizumab plus nab-paclitaxel could significantly improve the progression-free survival (PFS) of PD-L1-positive TNBC patients [7.5 months versus 5.0 months, hazard ratio = 0.62; 95% confidence interval (CI) = (0.49–0.78), p < 0.001]. Several studies also suggest a benefit of capecitabine-based chemotherapy in triple-negative BCBM patients.^17,18

Magnetic resonance imaging (MRI) is a sensitive imaging tool which is widely used in the detection, treatment planning, and follow-up of BCBM. Circulating tumor DNA (ctDNA) testing is promising regarding to tumor treatment response and progression.¹⁹ Derived from cell-free DNA (cfDNA), ctDNA analysis is a powerful surveillance tool for effective and continuous detection of potential tumor-related gene mutations.^20–23 Compared with imaging and serum biomarkers, ctDNA testing provides valuable and sensitive information about gene mutations in tumors following drug-based therapies.²⁴ ctDNA does provide valuable information, but ctDNA is less likely to be informative about brain disease than extra-cranial disease.²⁵ However, metastatic BC (MBC) is a systematic disease, and often times BC patients develop brain metastases after metastases have appeared systemically in the lung, liver, and/or bone.²⁶ Our previous study reports a ctDNA gene mutation profile in MBC patients.²⁴ Thymidine phosphorylase (TP53), PIK3CA, and ERBB2 are the top mutated genes. Angiogenesis-related vascular endothelial growth factor (VEGF) variation is not detected. Instead, fibroblast growth factor receptor 1 (FGFR1) variations are detected in 4/68 (5.88%) of MBC patients.²⁴ A total of 18% of BC patients are reported to be affected with FGFR aberrations, and the most common of which is FGFR1 amplification (~14%).²⁷ The FGFR pathway plays a major role in angiogenesis and metastasis; FGFR alterations promote carcinogenesis, cell proliferation, angiogenesis, and drug resistance.^28–30

In this study, we recruited 146 BCBM patients in our single center and retrospectively assessed the clinical risk factors. In addition, by using ctDNA testing, we examined the genetic risk factor in 195 single-center cases and in 1501 cases from Geneplus database.

Methods

BCBM patient cohort and data flow

The study was approved by the Ethics Committee at Hunan Cancer Hospital, the Affiliated Cancer Hospital of Xiangya School of Medicine/Central South University. As shown in Online Supplemental Figure S1 data flow, a total of 146 BCBM patients were enrolled in this study. Inclusion criteria were as follows: (a) CT/MRI confirmed BCBM; (b) tolerable to chemotherapy or target therapy with normal heart, liver, and renal function. Basic demographic and clinical information includes age of primary BC diagnosis, age of BCBM diagnosis, time-to-brain metastasis (TTB, from primary BC diagnosis to BCBM diagnosis), primary BC laterality, HR/HER2 status, TNM (T, primary tumor; N, nodal involvement; M, distant metastasis) stage of the primary BC at diagnosis, treatment history before BCBM diagnosis, treatment after BCBM diagnosis, and other metastatic sites.

Sample collection and plasma ctDNA testing

To investigate the genetic risk factor, plasma ctDNA test was performed. As shown in Figure S1, a total of 195 MBC patients received plasma ctDNA testing at the Department of Breast Cancer Medical Oncology in Hunan Cancer Hospital. Other inclusion criteria included: (a) ⩾1 measurable metastatic lesion; (b) age 18–70 years old; (c) normal liver and renal function (aspartate aminotransferase (AST) and alanine aminotransferase (ALT) > 0.5 upper limit of normal value (ULN), total bilirubin (TBIL) < 1.5 ULN, and creatinin (Cr) < 1.0 ULN); (d) routine-blood test: neutrophil count >2.0 × 10⁹/l, Hb >11 g/l, platelet count >100 × 10⁹/l. The exclusion criteria were: (a) serious heart disease; (b) multiple primary malignancies; (c) immune-deficiency; (d) organ-transplantation history. The plasma ctDNA testing study was also approved by the Ethics Committee of Hunan Cancer Hospital. Of these 195 patients, 30 were BCBM patients as described above, whereas the other 165 were non-brain-metastatic (BM) MBC patients. To compare the different genetic aberrations between BCBM patients and non-BM MBC patients, 1:1 propensity score matching (PSM) was performed to reduce the potential bias between these two subgroups. Propensity scores were calculated through logistic regression with covariates including age, treatment history, and metastatic disease (lung, liver, bone, and lymph node metastases) and recurrent disease. After PSM, 28 BCBM patients and 28 non-BM MBC patients were selected for genetic aberration comparison (heatmap).

From 2016, the Geneplus database pool collected 1501 BC samples; among these BC patients, 86 were BCBM patients. We collected the peripheral blood samples from the patients, and then extracted the DNA content according to the protocol.²⁴ We sequenced the genomic DNA (gDNA) and used the gDNA sequence as the control.

Target capture, next-generation sequencing, and data analysis

Sequencing libraries of ctDNA were prepared using the DNA Library Preparation Kit for Illumina (New England Biolabs, Ipswich, MA, USA). Custom biotinylated oligonucleotide probes (IDT, Coralville, IA, USA) covering the exons of 1021 genes that are highly mutated in 12 common solid tumors were used for hybrid capture, as described previously.³¹ The Illumina HiSeq 3000 Sequencing System (Illumina, San Diego, CA, USA) was used for DNA sequencing with a 2 × 101-bp paired-end strategy, as described previously.²⁴

Terminal adaptor sequences were removed from the raw sequencing data. Subsequently, reads with more than 50% low-quality bases, or more than 50% undefined bases, were discarded. The remaining reads were mapped to the reference human genome (hg19) using the Burrows-Wheel Aligner with default parameters. Picard’s Mark Duplicates tool (version 1.98) was used to identify duplicate reads. Local realignment and quality recalibration were performed using the Gene Analysis Toolkit. Single-nucleotide variants and small insertions and deletions were called using the MuTect2 algorithm, and further filtration and validation was performed according to established criteria.³¹ The Contra algorithm was used to identify somatic copy-number alterations defined using the ratio between the adjusted depths of ctDNA and control gDNA. After automatic calling, candidate variants were manually validated using an online visualization tool.

ctDNA gene aberration frequency

Total cfDNA included ctDNA and other normal cfDNA. Aberrations in ctDNA were identified by comparing the reference genome (hg19) and gDNA. The ctDNA aberration frequency was defined as the proportion of ctDNA gene aberrations in the total cfDNA.

Statistical analyses

Categorical variables were recorded as counts with percentage, and continuous variables were calculated as the mean with standard deviation and median with interquartile range. Chi-square tests were used to compare categorical variables between subgroups. When the sample size was less than five in the comparing subgroup, Fisher’s exact test were applied. Mentel–Haenszel chi-square tests were used when the number of compared subgroups was more than two. To compare continuous variables with symmetrical distributions across subgroups, an analysis of variance test was used. Both the univariate and multivariate Cox proportional hazards regression analyses were used to evaluate the prognostic factors for OS. Log rank tests and drawn Kaplan–Meier curves were also performed to assess the overall survival OS.

To search for significant ctDNA aberrations for BCBM, R package “ComplexHeatmap” was applied to rank the hot genetic aberrations in PSM BCBM and non-BM MBC subgroups. The top aberrations were put into a logistic model for BM prediction. A receiver operating characteristic (ROC) curve was drawn to determine the areas under the ROC curve (AUC) for BM. All statistical analyses were conducted by using SAS 9.4 or R 3.6.0 software. All tests of hypotheses were two-tailed and conducted at a significance level of 0.05 and at a marginal significant level of 0.15.

Results

Clinical risk factors for poor prognosis in BCBM patients

For the 146 recruited BCBM patients, the mean age at BC first diagnosis was 45.30 (±9.80) years old and the mean age at BCBM diagnosis was 48.48 (±9.64) years old. The mean TTB was 3.18 (±2.98) years (Online Supplemental Table S1). A total of 64.83% BCBM patients had bone metastasis and 76.56% BCBM patients had visceral metastasis. Both univariate and multivariate Cox regression analyses were performed to evaluate the potential clinical risk factors for poor OS in BCBM patients. Ki67 > 14% and metastatic brain tumors >3 were significant risk factor for OS in both univariate and multivariate Cox regression analyses (Table 1). Brain radiotherapy and chemotherapy after BM diagnosis were significant beneficial factors for longer OS in both univariate and multivariate Cox regression analyses. In multivariate Cox regression analysis, HER2-positive was a marginally significant favorable factor for OS (p = 0.10, Table 1). A Kaplan–Meier curve with Logrank test analyses showed that the median OS of all the enrolled 146 BCBM patients was approximately 31.8 months (Online Supplemental Figure S2a). The OS were longer in the patients with ⩽3 metastatic brain tumors than in the patients with >3 metastatic brain tumors (Logrank test p = 0.04, Online Supplemental Figure S2b). In the patients with Ki-67 ⩽ 14%, the OS were longer than in the patients with Ki-67 > 14% (Online Supplemental Figure S2c). After 18 months, the OS was longer in the patients with HER2 positive as compared with that of the patients with HER2 negative (Figure S2d).

Table 1.

Univariate and multivariate analysis of the clinical risk factors for OS.

Covariates	Level	Univariate		Multivariate
		Hazard ratio (95% CI)	p-value	Hazard ratio (95% CI)	p-value
Age at diagnosis (of BC)^*		1.016 (0.982, 1.052)	0.47	−	−
Age at brain metastasis (years)		1.006 (0.968, 1.045)	0.76	0.980 (0.869, 1.105)	0.74
TTB (years)^**		0.840 (0.713, 0.991)	0.04	1.185 (0.552, 2.545)	0.19
BC Laterality	Left	0.863 (0.405, 1.841)	0.70	0.615 (0.048, 7.883)	0.71
	Right	Ref
ER	Positive	0.751 (0.370, 1.523)	0.43	0.035 (0, 4.193)	0.17
	Negative	Ref		Ref
PR	Positive	0.692 (0.334, 1.431)	0.32	9.226 (0.276, 308.245)	0.21
	Negative	Ref		Ref
HER2	Positive	0.838 (0.503, 1.396)	0.50	0.497 (0.195, 1.267)	0.10
	Negative	Ref		Ref
HR/HER2 subtypes	Triple Negative	Ref		Ref
	HR+/HER2–	0.447 (0.177, 1.129)	0.09	0.540 (0.007, 41.856)	0.54
	HR–/HER2+	2.033 (0.618, 6.711)	0.24	1.318 (0.008, 142.857)	0.92
	HR+/HER2+	1.828 (0.646, 5.183)	0.26	1.113 (0.087,14.225)	0.95
Ki67 (%)	⩽14%	Ref		Ref
	>14%	6.842 (1.022, 45.805)	0.05	10.156 (1.009, 102.224)	0.05
Stage at BC diagnosis	I	Ref		Ref
	II	1.148 (0.316, 4.173)	0.83	0.352 (0.008, 16.537)	0.60
	III	0.895 (0.231, 3.471)	0.87	0.055 (0.001, 5.821)	0.22
	IV	1.658 (0.432, 6.362)	0.46	0.234 (0, 177.657)	0.67
Treatment history before BCBM diagnosis	Primary site surgery	0.775 (0.291, 2.065)	0.61	0.174 (0.002, 12.880)	0.43
	Primary site Radiation	1.330 (0.645, 2.742)	0.44	1.868 (0.109, 32.130)	0.32
	Chemotherapy	0.426 (0.057, 3.199)	0.41	0.866 (0.251, 3.091)	0.74
	Capecitabine usage^***	0.902 (0.439, 1.852)	0.78	0.382 (0.033, 4.477)	0.44
	Anti-Her2 targeting therapy	0.746 (0.460, 1.210)	0.25	0.805 (0.034, 23.676)	0.75
	Endocrine therapy	0.753 (0.385, 1.460)	0.52	0.917 (0.027, 49.464)	0.93
Treatment after BCBM diagnosis	Brain radiation	0.263 (0.113, 0.613)	0.002	0.150 (0.036, 0.625)	0.010
	Chemotherapy	0.286 (0.137, 0.597)	0.0009	0.014 (0.001, 0.702)	0.014
	Capecitabine usage^***	0.638 (0.272, 1.496)	0.30	2.434 (0.186, 31.644)	0.50
	Anti-Her2 targeting therapy	0.515 (0.179, 1.481)	0.22	0.708 (0.009, 54.457)	0.88
	Endocrine therapy	0.446 (0.106, 1.876)	0.27	0.502 (0.016, 54.457)	0.69
Brain metastases number	⩽3	Ref		Ref
	>3	2.208 (1.022, 4.768)	0.04	7.947 (1.012, 62.406)	0.05
Other metastasis^#	Bone	1.918 (0.821, 4.477)	0.13	1.765 (0.128, 24.401)	0.67
	Visceral	2.051 (0.825, 5.103)	0.12	4.685 (0.019, 242.439)	0.44

Age at diagnosis was evaluated by using univariate Cox regression; but this variable did not exist in the multivariate model because age at brain metastasis and TTB were analyzed in a multivariate model (Age at diagnosis=Age at brain metastasis-TTB).

TTB indicated the time from diagnosis of BC to the diagnosis time of brain metastasis.

***

Capecitabine usage represents the capecitabine usage history before or after BCBM diagnosis.

Other metastasis indicated the metastatic sites other than brain.

BC, breast cancer; BCBM, breast cancer and brain metastasis; CI, confidence interval; ER, estrogen receptor; HER2, human epidermal growth factor receptor-2; PR, progesterone receptor; TTB, time-to-brain metastasis.

Genetic aberration of BCBM patients

BCBM group and non-BM MBC group had distinct ctDNA aberration patterns (Figure 1). BCBM patients had a significantly higher FGFR aberration rate compared with non-BM MBC patients (32.14% versus 3.57%, Fisher’s Exact test p = 0.01). FLT1 gene aberration was also higher in BCBM patients, while ALK, CCNE1 and CDK12 aberrations were more frequent in non-BM MBC patients.

Figure 1.

Circulating tumor DNA gene mutation profiles for matched 28 brain metastatic breast cancer (BC) patients (a) and 28 non-brain-metastatic metastatic BC patients (b).

Compared with non-BM MBC patients, more BCBM patients were HER2-positive (40% versus 23.75%, p = 0.06, Table S2), and more BCBM patients had bone or liver metastases (p = 0.04 and p = 0.001, respectively). Based on these features, a panel of genetic events (including aberrant FGFR, TP53, FLT1, and wild-type ALK, CDK12 and CCNE1) and HER2-positive were correlated with an increased risk of BM (AUC = 0.8050, Online Supplemental Figure S3a). An even smaller panel of genetic aberrations (FGFR, TP53, FLT1) and HER2-positive were still significantly correlated with an increased risk of BM (AUC = 0.7713). But such a panel of genetic and HER2-positive events was less likely correlated with liver, lung, bone, and lymph node metastases (AUC: 0.6136 ~ 0.7075, Online Supplemental Figure S3b–e).

Table 2.

Differences of metastases/recurrence between FGFR-aberrant group and wild-type FGFR group.

	FGFR group (n = 22)	wild type-FGFR group (n = 173)	p-value^*
Brain	9 (40.91%)	12 (6.94%)	<0.0001
Liver	12 (54.55%)	36 (20.81%)	0.0005
Lung	12 (54.55%)	42 (24.28%)	0.003
Bone	12 (54.55%)	59 (34.10%)	0.06
Lymph node	15 (68.18%)	70 (40.46%)	0.01
Recurrence	40 (23.12%)	8 (36.36%)	0.17

p-values were calculated by using Chi-square tests or Fisher’s exact tests (n < 5) for categorical variables.

FGFR aberration in MBC patients

FGFR pathway plays a major role in angiogenesis, metastasis, cell proliferation and drug resistance.^28–30 In squamous cell lung cancer and BC, FGFR inhibitors showed promising antitumor activity, especially for patients with FGFR1 and/or FGFR3 amplification.³² Here, among 195 MBC patients, FGFR aberration was not only significantly related to poor PFS (Logrank test p = 0.029, Figure 2), but also had AUC of 0.6790 in predicting brain metastases (Online Supplemental Figure S3b–e), suggesting a strong correlation of FGFR aberration with poor prognosis and brain metastases. Compared with non-FGFR aberrant group, FGFR-aberrant group had significantly more brain, liver, lung, and lymph node metastatic patients (p < 0.05, Table 2).

Figure 2.

Kaplan–Meier curves for progression-free survival (PFS) probabilities (probs.) stratified by circulating tumor DNA FGFR aberrations.

Among 197 MBC patients, 22 patients had FGFR aberrations, including 14 FGFR1 aberrations, 4 FGFR2 aberrations, 3 FGFR3 aberrations and 1 FGFR4 aberration (Figure 3). Besides TP53 and PIK3CA, MAP3K1 mutation was the most frequent in FGFR-aberrant patients (Figure 3 and Table 3, p = 0.001). A small panel of genetic aberrations (including TP53, PIK3CA, MAP3K1, ESR1, and DNMT3A) could potently predict the FGFR aberration in MBC patients (AUC = 0.7349, Online Supplemental Figure S3f). Our findings suggested that FGFR aberration was coupled with multiple oncogenic pathways and drug-resistant pathways, such MAPK/ERK signaling pathway and mTOR pathway.³³

Figure 3.

Circulating tumor DNA gene mutation profiles (top) and metastases (bottom) for 22 FGFR-aberrant metastatic breast cancer patients.

Table 3.

Differences of ctDNA mutation pattern between FGFR-aberrant group and wild-type group.

	FGFR group (n = 22)	wild type-FGFR group (n = 173)	p-value^*
TP53	12 (54.55%)	84 (48.55%)	0.60
PIK3CA	11 (50.00%)	62 (35.84%)	0.20
MAP3K1	5 (22.73%)	4 (2.31%)	0.001
ESR1	3 (13.64%)	13 (7.51%)	0.40
DNMT3A	3 (13.64%)	15 (8.67%)	0.43

p-values were calculated by using Chi-square tests or Fisher’s exact tests (n < 5) for categorical variables.

FGFR1 aberration increased in MBC patients with brain metastases

FGFR1 aberration was the most frequent FGFR aberration in BCBM patients (Figure 3). To further explore the FGFR1 aberration pattern in MBC patients, we used ctDNA surveillance data from Geneplus database. Among 1501 BC patients in Geneplus database (Table 4), 379 patients had bone metastases but the FGFR1 aberration rate was almost the same with non-bone metastatic BC patients (4.75% versus 3.48%, p = 0.26). In lung metastatic BC patients, the FGFR1 aberration was significantly higher than the rate in non-lung metastatic BC patients (5.83% versus 3.26%, p = 0.04). For liver metastatic and brain metastatic BC patients, the FGFR1 aberration rates were even higher (8.18% and 8.14%, respectively), significantly higher than those without liver- and brain-metastases (p < 0.0001 and p = 0.03, respectively, Table 4). Based on these findings, we hypothesized that patients with more aggressive metastases would have higher FGFR1 aberration rate.

Table 4.

FGFR1 aberration rate comparison in bone-, brain-, lung-, and liver-metastatic breast cancer patients in Geneplus database.

Metastatic sites	Level	Number	FGFR1 aberration		p-value^*
			Yes	No
Bone	Yes	379	18 (4.75%)	361 (95.25%)	0.26
	No	1122	39 (3.48%)	1083 (96.52%)
Lung	Yes	309	18 (5.83%)	291 (94.17%)	0.04
	No	1195	39 (3.26%)	1156 (96.74%)
Liver	Yes	269	22 (8.18%)	247 (91.82%)	<0.0001
	No	1232	35 (2.84%)	1197 (97.16%)
Brain	Yes	86	7 (8.14%)	79 (91.86%)	0.03
	No	1415	50 (3.53%)	1365 (96.47%)

p-values were calculated by using Chi-square tests or Fisher’s exact tests (n < 5) for categorical variables.

Among 57 FGFR1-aberrant BC patients from Geneplus database, 61 FGFR1 aberrations were identified. More than half were FGFR1 amplification (35/61 [57.38%], Online Supplemental Table S3), one (1.64%) was splicing variation, three (4.92%) were stop-gain mutations, and the rest twenty-two (36.07%) were missense mutations. Among seven FGFR1-aberrant BCBM patients from Geneplus database, eight FGFR1 aberrations were identified. Five (62.5%) were copy number amplification, one (12.5%) was stop-gain mutation and two (25.0%) were missense mutations. Among 14 FGFR1 aberrant BCBM patients (Figure 3), ten were amplification, one was stop-gain mutation and three were missense mutations. The average level of FGFR1 amplification was 4.05 (±2.32) copy number. The median level of FGFR1 amplification was 3.2 [range = (2.1, 10)] copy number.

FGFR aberrations predict poor PFS

Kaplan–Meier curve with lifetest showed that FGFR aberrations were also significantly related to poor PFS (Logrank test p = 0.029, Figure 2), suggesting a strong correlation of FGFR aberration with poor treatment response. The median PFS for FGFR-wild type patients and FGFR-aberrant patients was 7.5 (95% CI = 5.3–9.3) months and 4.7 (95% CI = 2.0–7.1) months, respectively. As shown in Table 5, 13/18 (72%) FGFR aberrant MBC patients progressed within 6 months. 60% FGFR1-amplified patients progressed within 3 months. Only one FGFR1 amplified HR+/HER2+ patient (ID53) with capecitabine + lapatinib + Herceptin treatment had a PFS longer than 6 months, and the ctDNA surveillance data also suggested a decrease in tumor mutation burden. The treatment response for FGFR1 stop-gain and missense mutant patients (ID52, 22, 104 and 177) was better than FGFR1-amplified patients, with a longer PFS (patients with PFS > 6 months were highlighted in red). Besides FGFR1 amplification, FGFR2 and FGFR3 aberrant patients also had short PFS, except one FGFR2-amplified patient with Exemestane plus Chidamide treatment (ID58). Consistent to the RECIST evaluation, ctDNA surveillance showed FGFR aberration sustained or increased in resistant patients (PFS < 6 months) and FGFR aberration decreased in sensitive patients (PFS > 6 months, Table 5).

Table 5.

Treatment response for BC patients with FGFR aberrations.

ID	HR/HER2 subtype	BM	Treatment	Genes	Aberrations	Level of amplification^#	ctDNA surveillance^## for FGFR aberrations	Progression^*	PFS^**	COSMIC record^Ϯ
72	TNBC	No	NVB + platin	FGFR1	Amplification	3	Stable	1	2.0	FGFR1 amplification (recorded in COSMIC for 910 times) was detected in breast cancer, pancreatic cancer, and lung cancer. FGFR1 amplification could activate downstream signaling pathways, including FGFR signaling pathway, PIP3-activated AKT signaling pathway, etc. In vitro studies showed that lapatinib was able to effectively inhibit the proliferation of FGFR1-amplified lung cancer and breast cancer cell lines.³⁴ FGFR1 amplification also related to drug resistance to endocrine therapy.^35–37
53	HR+/HER2+	Yes	Capecitabine + lapatinib + Herceptin	FGFR1	Amplification	10	Decrease	1	7.1
117	HR+/HER2–	No	Goserelin + exemestane	FGFR1	amplification	2.4	Increase	1	2.3
175	HR+/HER2–	Yes	Fulvestrant	FGFR1	amplification	5.3	Increase	1	2.0
144	HR+/HER2–	No	Palbociclib + fulvestrant	FGFR1	amplification	2.8	Increase	1	2.0
125	HR+/HER2–	No	NVB + capecitabine	FGFR1	amplification	3.2	Increase	1	3.5
145	TNBC	Yes	Abraxane	FGFR1	amplification	2.1	Increase	1	2.0
190	HR+/HER2–	No	Taxol	FGFR1	amplification	3.6	Increase	1	3.2
130	HR+/HER2+	No	Lapatinib + capecitabine	FGFR1	Amplification	4.9	Stable	0	3.0 censored
16	TNBC	No	Doxorubicin + Taxol	FGFR1	amplification	3.2	Stable	1	2.0
52	HR+/HER2–	Yes	NVB + Xeloda	FGFR1	p.R809^*		Decrease	1	10.4	Pathogenic (score 0.91), reported in cholangiocarcinoma, chondrobla stoma and colon cancer³⁸
22	HR+/HER2–	Yes	Taxol + cisplatin	FGFR1	p.G459R		Stable	0	3.0 censored	No record in COSMIC
104	HR+/HER2–	No	Everolimus + exemestane	FGFR1	p.N546K		Decrease	1	7.9	Pathogenic (score 0.95). p.N546K mutation located in the kinase domain, facilitated the phosphorylation of FGFRs and FRS2D, and activated RAS-MAPK pathway^39,40
177	HR+/HER2–	No	Abemaciclib + fulvestrant	FGFR1	p.N546K		Decrease	0	12.8
58	HR+/HER2–	Yes	Exemestane + chidamide	FGFR2	amplification	10	Decrease	1	7.5	FGFR2 amplification related to tumorigenesis and metastasis.
142	HR+/HER2–	No	Epirubicin + CTX + Taxol	FGFR2	amplification	20.3	Increase	1	4.8
29	TNBC	No	Doxorubicin + Taxol	FGFR2	p.V463D		Increase	1	4.7	No record in COSMIC
66	HR+/HER2–	No	Taxol + capecitabine	FGFR2	p.Y375C		Stable	1	5.6	FGFR2 p.Y375C activate downstream MAPK signaling pathway, and induce malignant proliferation^41,42
80	HR–/HER2+	Yes	NVB + lapatinib + everolimus	FGFR3-TACC3	fusion		Stable	1	4.1	In lung adenocarcinoma, bladder cancer and glioma, FGFR3-TACC3 induced malignant proliferation.⁴³
119	TNBC	No	Doxorubicin + CTX + Taxol	FGFR3	p.l16_A18dup		Stable	0	1.0 censored	No record in COSMIC
186	HR+/HER2–	Yes	Taxol + capecitabine	FGFR3	p.R571W		Increase	1	1.2	No record in COSMIC
11	HR+/HER2–	Yes	Radiation + gemcitabine	FGFR4	p.R3128		Decrease	1	6.3	No record in COSMIC

Patients with PFS > 6months were highlighted in red.

Level of amplification indicated the copy number of FGFR1/2 gene amplifications.

ctDNA surveillance represented the change trend of aberrant levels of FGFR genes in response of treatment, detected by ctDNA testing. Increase indicated an increase of FGFR aberration during treatment; decrease indicated the decreasing trend of FGFR aberration during treatment.

Progression indicated whether patient progressed (1) or not (0) till the date of last follow-up.

PFS represented the progression-free survival (months) for progressed patients (progression = 1), or the duration to the date of last follow-up for censored patients (progression = 0).

COSMIC record represented the pathological effect of FGFR mutations that were recorded in COSMIC.

BC, breast cancer; BM, brain metastasis; COSMIC, Catalogue Of Somatic Mutations In Cancer; ctDNA, circulating tumor DNA; HER2, human epidermal growth factor receptor-2; HR+/-, hormone receptor positive or negative; PFS, progression-free survival; TNBC, triple-negative breast cancer.

Specifically, FGFR1 missense mutant HR-positive MBC patient had longer PFS than FGFR1 amplified HR-positive MBC patient. ID144 and ID177 were both HR-positive MBC patients with CDK4/6 inhibitor plus Fulvestrant treatment. ID144 had FGFR1 amplification and progressed within 2 months of treatment. ID177 had FGFR1 p.N546K missense mutation, and she kept partial response for more than 1 year (Table 4). Everolimus plus exemetane also showed potency in FGFR1 p.N546K missense mutant HR-positive patient (ID104). Compared to epirubicin plus cyclophosphamide plus taxel (ECT) regimen (ID142), chidamide plus exemestane was more effective in FGFR2-amplified HR-positive patient (ID58). By using the ctDNA test, we could find more specific drug-resistant mutations, such as ESR1, PIK3CA p.E545K/E525K mutations, FGFR1 amplification, etc.

FGFR aberrations concentrated in HR-positive BCBM patients

Compared to non-BM MBC patients, more BCBM patients were HER2-positive (40% versus 23.75%, p = 0.06, Table S2). But compared with FGFR-wild type patients, more FGFR-aberrant patients were ER-positive or PR-positive (ER: 68.18% versus 48.55%, p = 0.08; PR: 68.18% versus 43.93%, p = 0.03; Table S4). Both FGFR aberration and HER2-positive were correlated with BM, but their combination showed a more significant correlation. As shown in Online Supplemental Figure S3a, the combination of FGFR aberration and HER2-positive was significantly correlated with the increased risk of BM (AUC = 0.7228), but such a correlation of individual aberration was not as significant as their combination (FGFR aberration alone AUC = 0.6790, HER2-positive alone AUC = 0.5431, Figure S3A).

Discussion

In practice, we found that some other brain metastases sequentially occurred after bone or other visceral metastases, especially in HR-positive BC patients, suggesting brain metastases as the final deadly stage for this group of patients; some brain metastases appeared earlier than bone and other visceral metastases, especially in HER2-positive BC patients. In this study, we found that FGFR aberrations concentrated in HR-positive patients (Online Supplemental Table S4); the FGFR aberrant group also had significantly more liver, lung, and brain metastases than the non-FGFR aberrant group (Table 3), implying a role of FGFR aberration in the late stage (migration and metastases) of HR-positive patients (especially in luminal B subtype Ki > 14%). HER2 was another important factor related to BM: the proportion of HER2-positive tumors was higher in BCBM patients, compared with non-BM MBC patients (Online Supplemental Table S2). The combination of HER2-positive and FGFR aberration could reach an AUC of 72.28% in the prediction of brain metastases (Online Supplemental Figure S3a), indicating that 72.28% BCBMs were either HER2-positive and/or FGFR-aberrant.

Genetic risk factor FGFR aberration was identified to relate to BCBM in both our single-center and Geneplus database. Similarly, in the brain metastases of lung adenocarcinomas, FGFR1 amplification is also reported to be higher than the expected rate.⁴⁴ FGFR1 amplification is identified in up to 10% of all BCs, 16–27% of luminal B type BCs, and is correlated with endocrine therapy resistance.^35,36 In HER2-positive BCs, the co-amplification of FGFR1, ERBB1, and ERBB2 results in a worse prognosis compared with patients with individual amplification or without amplification.⁴⁵ In this study, among 22 detected FGFR-aberrant patients, 16 were HR-positive MBC patients (14 were HR+/HER2- and 2 were HR+/HER2+), 5 were TNBC and 1 was HR–/HER2+ (Table 5). Moreover, ctDNA FGFR aberrations were more frequently detected in ER-positive [15/99 (15.15%) versus 7/96 (7.29%), p = 0.08] or PR-positive [15/91 (16.48%) versus 7/104 (6.73%), p = 0.03, Online Supplemental Table S2] MBC patients than ER-negative or PR-negative MBC patients. FGFR aberrant HR+/HER2- patients had short PFS and poor drug response (Table 5), suggesting FGFR aberrations might result in therapy resistance and tumor progression in HR-positive MBC patients. Among 22 detected FGFR-aberrant patients, 5 mutations were not ones that were known to be pathogenic (no record in Catalogue Of Somatic Mutations In Cancer), including FGFR1 p.G459R missense mutation, FGFR2 p.V463D missense mutation, FGFR3 p.l16_A18dup frameshift mutation, FGFR3 p.R571W missense mutation, and FGFR4 p.R3128 missense mutation (Table 5). Some patients with these non-pathogenic mutations had poor response to anti-cancer treatment, so in future we need to focus on the function of these mutations.

Recently, due to the application of new targeting agents, such as pertuzumab, ado-trastuzumab emtansine, and lapatinib, the survival outcome has been improved for HER2-positive BC patients with brain metastases. Anti-HER2 therapies could extend the median OS from a few months up to 16–20 months.^46,47 Lapatinib plus capecitabine regimen is, at present, the most effective option for HER2-postive BCBM patients, with favorable tumor response, disease control, and survival rates.^10,48,49 Lapatinib is a lipophilic, small-molecular inhibitor of EGFR and HER2.⁵⁰ For patient ID53 in this study, we found that the brain metastatic lesion shrank after capecitabine plus lapatinib and trastuzumab treatment. Besides lapatinib, a recent study suggests neratinib plus capecitabine is active against refractory HER2-positive BCBM⁵¹; Tucatinib in combination with capecitabine and trastuzumab also shows preliminary anti-tumor activity in advanced HER2-positive BCBMs⁴⁸ (HER2Climb trial to be presented at the San Antonio Breast Cancer Symposium 2019). In addition, pyrotinib plus capecitabine exhibits efficacy in HER2-positive BCBM (Phoenix TRAIL, 2019 American Society of Clinical Oncology).

Besides FGFR, TP overexpression correlates with the pro-angiogenic (VEGF) to promote angiogenesis and metastasis in human tumor tissues.^52,53 Since capecitabine is a prodrug which is activated to 5-fluorocytidine by carboxylesterase, cytidine-deaminase and TP, capecitabine is supposed to be more effective in TP-positive tumors, including BC.⁵² In this study, capecitabine-based treatment showed activity in FGFR-aberrant BCBM patients (ID52 and 53), suggesting a potential crosstalk of TP and FGFR in BCBM. In HER2-positive BCBM patients, lapatinib (small molecule tyrosine kinase inhibitor) is particularly active when combining with capecitabine.⁵⁴ By using the clinical and genetic features, we successfully constructed BM prediction models. Both FGFR aberration and HER2-positive were important elements in the prediction model.

Novelty and impact statement

A group of genetic events could forecast the occurrence of brain metastases in BCs (BCBM), including FGFR, TP53 and FLT1 genetic aberrations, and HER2-positive. FGFR aberration alone not only increased BCBM, but also was associated with poor prognosis in metastatic BCs (MBC). FGFR aberration, especially FGFR1 amplification, correlated with short progression-free survival in MBC patients.

Supplemental Material

Table_S1 – Supplemental material for FGFR aberrations increase the risk of brain metastases and predict poor prognosis in metastatic breast cancer patients

Supplemental material, Table_S1 for FGFR aberrations increase the risk of brain metastases and predict poor prognosis in metastatic breast cancer patients by Ning Xie, Can Tian, Hui Wu, Xiaohong Yang, Liping liu, Jing Li, Huawu Xiao, Jianxiang Gao, Jun Lu, Xuming Hu, Min Cao, Zhengrong Shui, Yu Tang, Xiao Wang, Jianbo Yang, Zhe-Yu Hu and Quchang Ouyang in Therapeutic Advances in Medical Oncology

Supplemental Material

Table_S2 – Supplemental material for FGFR aberrations increase the risk of brain metastases and predict poor prognosis in metastatic breast cancer patients

Supplemental material, Table_S2 for FGFR aberrations increase the risk of brain metastases and predict poor prognosis in metastatic breast cancer patients by Ning Xie, Can Tian, Hui Wu, Xiaohong Yang, Liping liu, Jing Li, Huawu Xiao, Jianxiang Gao, Jun Lu, Xuming Hu, Min Cao, Zhengrong Shui, Yu Tang, Xiao Wang, Jianbo Yang, Zhe-Yu Hu and Quchang Ouyang in Therapeutic Advances in Medical Oncology

Supplemental Material

Table_S3 – Supplemental material for FGFR aberrations increase the risk of brain metastases and predict poor prognosis in metastatic breast cancer patients

Supplemental material, Table_S3 for FGFR aberrations increase the risk of brain metastases and predict poor prognosis in metastatic breast cancer patients by Ning Xie, Can Tian, Hui Wu, Xiaohong Yang, Liping liu, Jing Li, Huawu Xiao, Jianxiang Gao, Jun Lu, Xuming Hu, Min Cao, Zhengrong Shui, Yu Tang, Xiao Wang, Jianbo Yang, Zhe-Yu Hu and Quchang Ouyang in Therapeutic Advances in Medical Oncology

Supplemental Material

Table_S4_FGFR – Supplemental material for FGFR aberrations increase the risk of brain metastases and predict poor prognosis in metastatic breast cancer patients

Supplemental material, Table_S4_FGFR for FGFR aberrations increase the risk of brain metastases and predict poor prognosis in metastatic breast cancer patients by Ning Xie, Can Tian, Hui Wu, Xiaohong Yang, Liping liu, Jing Li, Huawu Xiao, Jianxiang Gao, Jun Lu, Xuming Hu, Min Cao, Zhengrong Shui, Yu Tang, Xiao Wang, Jianbo Yang, Zhe-Yu Hu and Quchang Ouyang in Therapeutic Advances in Medical Oncology

Supplemental Material

Table_S5-brain – Supplemental material for FGFR aberrations increase the risk of brain metastases and predict poor prognosis in metastatic breast cancer patients

Supplemental material, Table_S5-brain for FGFR aberrations increase the risk of brain metastases and predict poor prognosis in metastatic breast cancer patients by Ning Xie, Can Tian, Hui Wu, Xiaohong Yang, Liping liu, Jing Li, Huawu Xiao, Jianxiang Gao, Jun Lu, Xuming Hu, Min Cao, Zhengrong Shui, Yu Tang, Xiao Wang, Jianbo Yang, Zhe-Yu Hu and Quchang Ouyang in Therapeutic Advances in Medical Oncology

Footnotes

Acknowledgements

We would like to thank the researchers from Geneplus Beijing Institute as these individuals sequenced all patients’ ctDNA and gDNA samples for this study.

Author note

Jianbo Yang is also affiliated with Fujian Medical University Union Hospital, Fuzhou, China.

Author contributions

Dr. Quchang Ouyang and Zhe-Yu Hu have the full access to all data in the study and take responsibility for the integrity and accuracy of the data analysis. Study concept and design: Zhe-Yu Hu, Quchang Ouyang. Data acquisition: All authors. Data analysis and interpretation: Zhe-Yu Hu and Quchang Ouyang. Drafting of the manuscript: Zhe-Yu Hu. Critical revision of the manuscript for important intellectual content: All authors

Conflict of interest statement

The authors declare that there is no conflict of interest.

Ethics approval and consent to participate

The study was approved by the Ethics Committee at Hunan Cancer Hospital, the Affiliated Cancer Hospital of Xiangya School of Medicine/Central South University. The ethical board approval number is 2018119. The authors declare that they have obtained ethical approval and patients consent to participate.

Funding

The authors disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: Hunan Provincial Natural Science Youth Foundation Project Grant No.2019JJ50356 (Recipient: Zhe-Yu Hu); Hunan Provincial Health and Sanitation Committee Project Grant No. C2019070 (Recipient: Zhe-Yu Hu); Key Research Program in Hunan Social Development Grant No. 2019SK2032 (Recipient: Zhe-Yu Hu); Changsha Science and Technology Project Program No.30749 (Recipient: Zhe-Yu Hu).

ORCID iD

Zhe-Yu Hu

Statement about data availability

We used the data from Geneplus database and from the Hunan Cancer Hospital. Geneplus database is not a public database. Its data will be made available upon reasonable request. ctDNA data from Hunan Cancer Hospital is available in supplemental Table S5 (BM patients) and (non-BM MBC patients).

Supplemental material

Supplemental material for this article is available online.

References

Miller

Weathers

Haney

, et al Occult central nervous system involvement in patients with metastatic breast cancer: prevalence, predictive factors and impact on overall survival. Ann Oncol 2003; 14: 1072–1077.

Lin

Bellon

Winer

. CNS metastases in breast cancer. J Clin Oncol 2004; 22: 3608–3617.

Lin

Claus

Sohl

, et al Sites of distant recurrence and clinical outcomes in patients with metastatic triple-negative breast cancer: high incidence of central nervous system metastases. Cancer 2008; 113: 2638–2645.

Brogi

Murphy

Johnson

, et al Breast carcinoma with brain metastases: clinical analysis and immunoprofile on tissue microarrays. Ann Oncol 2011; 22: 2597–2603.

Wang-Gillam

Bodoky

, et al Nanoliposomal irinotecan with fluorouracil and folinic acid in metastatic pancreatic cancer after previous gemcitabine-based therapy (NAPOLI-1): a global, randomised, open-label, phase 3 trial. Lancet 2016; 387: 545–557.

Zhang

Chen

, et al Systemic treatment after whole-brain radiotherapy may improve survival in RPA class II/III breast cancer patients with brain metastasis. J Neurooncol 2013; 114: 181–189.

Rostami

Mittal

Rostami

, et al Brain metastasis in breast cancer: a comprehensive literature review. J Neurooncol 2016; 127: 407–414.

Bartsch

Berghoff

Pluschnig

, et al Impact of anti-HER2 therapy on overall survival in HER2-overexpressing breast cancer patients with brain metastases. Br J Cancer 2012; 106: 25–31.

Sutherland

Ashley

Miles

, et al Treatment of HER2-positive metastatic breast cancer with lapatinib and capecitabine in the lapatinib expanded access programme, including efficacy in brain metastases–the UK experience. Br J Cancer 2010; 102: 995–1002.

10.

Metro

Foglietta

Russillo

, et al Clinical outcome of patients with brain metastases from HER2-positive breast cancer treated with lapatinib and capecitabine. Ann Oncol 2011; 22: 625–630.

11.

Lee

Ledermann

Kohn

. PARP Inhibitors for BRCA1/2 mutation-associated and BRCA-like malignancies. Ann Oncol 2014; 25: 32–40.

12.

Venur

Leone

. Targeted therapies for brain metastases from breast cancer. Int J Mol Sci 2016; 17: 1543.

13.

Leone

. Breast cancer brain metastases: the last frontier. Exp Hematol Oncol 2015; 4: 33.

14.

Ramkissoon

Xie

, et al Combination inhibition of PI3K and mTORC1 yields durable remissions in mice bearing orthotopic patient-derived xenografts of HER2-positive breast cancer brain metastases. Nat Med 2016; 22: 723–726.

15.

Shonka

Venur

Ahluwalia

. Targeted treatment of brain metastases. Curr Neurol Neurosci Rep 2017; 17: 37.

16.

Schmid

Adams

Rugo

, et al Atezolizumab and nab-paclitaxel in advanced triple-negative breast cancer. N Engl J Med 2018; 379: 2108–2121.

17.

Lee

, et al Irinotecan and capecitabine combination chemotherapy in a patient with triple-negative breast cancer relapsed after adjuvant chemotherapy with anthracycline and taxane. Tumori 2015; 101: e9–e12.

18.

Liu

, et al Apatinib + CPT-11 + S-1 for treatment of refractory brain metastases in patient with triple-negative breast cancer: case report and literature review. Medicine (Baltimore) 2018; 97: e0349.

19.

Franchino

Ruda

Soffietti

. Mechanisms and therapy for cancer metastasis to the brain. Front Oncol 2018; 8: 161.

20.

Cheng

Liu

Jiang

, et al Analysis of ctDNA to predict prognosis and monitor treatment responses in metastatic pancreatic cancer patients. Int J Cancer 2017; 140: 2344–2350.

21.

Guibert

Mazieres

Delaunay

, et al Monitoring of KRAS-mutated ctDNA to discriminate pseudo-progression from true progression during anti-PD-1 treatment of lung adenocarcinoma. Oncotarget 2017; 8: 38056–38060.

22.

Tie

Kinde

Wang

, et al Circulating tumor DNA as an early marker of therapeutic response in patients with metastatic colorectal cancer. Ann Oncol 2015; 26: 1715–1722.

23.

Khan

Rata

Cunningham

, et al Functional imaging and circulating biomarkers of response to regorafenib in treatment-refractory metastatic colorectal cancer patients in a prospective phase II study. Gut. Epub ahead of print 8 August 2017 DOI: 10.1136/gutjnl-2017-314178.

24.

Xie

Tian

, et al Identifying circulating tumor DNA mutation profiles in metastatic breast cancer patients with multiline resistance. EBioMedicine 2018; 32: 111–118.

25.

Bettegowda

Sausen

Leary

, et al Detection of circulating tumor DNA in early- and late-stage human malignancies. Sci Transl Med 2014; 6: 224ra224.

26.

Weil

Palmieri

Bronder

, et al Breast cancer metastasis to the central nervous system. Am J Pathol 2005; 167: 913–920.

27.

Helsten

Elkin

Arthur

, et al The FGFR landscape in cancer: analysis of 4,853 tumors by next-generation sequencing. Clin Cancer Res 2016; 22: 259–267.

28.

Lieu

Heymach

Overman

, et al Beyond VEGF: inhibition of the fibroblast growth factor pathway and antiangiogenesis. Clin Cancer Res 2011; 17: 6130–6139.

29.

Hierro

Rodon

Tabernero

. Fibroblast growth factor (FGF) receptor/FGF inhibitors: novel targets and strategies for optimization of response of solid tumors. Semin Oncol 2015; 42: 801–819.

30.

Wei

Yin

, et al Inhibition of FGFR signaling by PD173074 improves antitumor immunity and impairs breast cancer metastasis. Breast Cancer Res Treat 2014; 143: 435–446.

31.

Yang

Chu

Zhang

, et al Technical validation of a next-generation sequencing assay for detecting clinically relevant levels of breast cancer-related single-nucleotide variants and copy number variants using simulated cell-free DNA. J Mol Diagn 2017; 19: 525–536.

32.

Dienstmann

Rodon

Prat

, et al Genomic aberrations in the FGFR pathway: opportunities for targeted therapies in solid tumors. Ann Oncol 2014; 25: 552–563.

33.

Babina

Turner

. Advances and challenges in targeting FGFR signalling in cancer. Nat Rev Cancer 2017; 17: 318–332.

34.

Gozgit

Wong

Moran

, et al Ponatinib (AP24534), a multitargeted pan-FGFR inhibitor with activity in multiple FGFR-amplified or mutated cancer models. Mol Cancer Ther 2012; 11: 690–699.

35.

Elbauomy Elsheikh

Green

Lambros

, et al FGFR1 amplification in breast carcinomas: a chromogenic in situ hybridisation analysis. Breast Cancer Res 2007; 9: R23.

36.

Turner

Pearson

Sharpe

, et al FGFR1 amplification drives endocrine therapy resistance and is a therapeutic target in breast cancer. Cancer Res 2010; 70: 2085–2094.

37.

Balko

Mayer

Sanders

, et al Discordant cellular response to presurgical letrozole in bilateral synchronous ER+ breast cancers with a KRAS mutation or FGFR1 gene amplification. Mol Cancer Ther 2012; 11: 2301–2305.

38.

Zehir

Benayed

Shah

, et al Mutational landscape of metastatic cancer revealed from prospective clinical sequencing of 10,000 patients. Nat Med 2017; 23: 703–713.

39.

Bennett

Tan

Alcantara

, et al Mosaic activating mutations in FGFR1 cause encephalocraniocutaneous lipomatosis. Am J Hum Genet 2016; 98: 579–587.

40.

Rivera

Gayden

Carrot-Zhang

, et al Germline and somatic FGFR1 abnormalities in dysembryoplastic neuroepithelial tumors. Acta Neuropathol 2016; 131: 847–863.

41.

Pollock

Gartside

Dejeza

, et al Frequent activating FGFR2 mutations in endometrial carcinomas parallel germline mutations associated with craniosynostosis and skeletal dysplasia syndromes. Oncogene 2007; 26: 7158–7162.

42.

Wang

Zhou

Oberoi

, et al p38 Inhibition ameliorates skin and skull abnormalities in Fgfr2 Beare-Stevenson mice. J Clin Invest 2012; 122: 2153–2164.

43.

Singh

Chan

Zoppoli

, et al Transforming fusions of FGFR and TACC genes in human glioblastoma. Science 2012; 337: 1231–1235.

44.

Preusser

Berghoff

Berger

, et al High rate of FGFR1 amplifications in brain metastases of squamous and non-squamous lung cancer. Lung Cancer 2014; 83: 83–89.

45.

Chen

Qiu

Guo

, et al FGFR1 and HER1 or HER2 co-amplification in breast cancer indicate poor prognosis. Oncol Lett 2018; 15: 8206–8214.

46.

Leyland-Jones

. Human epidermal growth factor receptor 2-positive breast cancer and central nervous system metastases. J Clin Oncol 2009; 27: 5278–5286.

47.

Brufsky

Mayer

Rugo

, et al Central nervous system metastases in patients with HER2-positive metastatic breast cancer: incidence, treatment, and survival in patients from registHER. Clin Cancer Res 2011; 17: 4834–4843.

48.

Murthy

Borges

Conlin

, et al Tucatinib with capecitabine and trastuzumab in advanced HER2-positive metastatic breast cancer with and without brain metastases: a non-randomised, open-label, phase 1b study. Lancet Oncol 2018; 19: 880–888.

49.

Petrelli

Ghidini

Lonati

, et al The efficacy of lapatinib and capecitabine in HER-2 positive breast cancer with brain metastases: a systematic review and pooled analysis. Eur J Cancer 2017; 84: 141–148.

50.

Ryan

Ibrahim

Cohen

, et al FDA drug approval summary: lapatinib in combination with capecitabine for previously treated metastatic breast cancer that overexpresses HER-2. Oncologist 2008; 13: 1114–1119.

51.

Freedman

Gelman

Anders

, et al TBCRC 022: A phase II trial of neratinib and capecitabine for patients with human epidermal growth factor receptor 2-positive breast cancer and brain metastases. J Clin Oncol 2019; 37: 1081–1089.

52.

Bonotto

Bozza

Di Loreto

, et al

Making capecitabine targeted therapy for breast cancer: which is the role of thymidine phosphorylase?

Clin Breast Cancer 2013; 13: 167–172.

53.

Relf

LeJeune

Scott

, et al Expression of the angiogenic factors vascular endothelial cell growth factor, acidic and basic fibroblast growth factor, tumor growth factor beta-1, platelet-derived endothelial cell growth factor, placenta growth factor, and pleiotrophin in human primary breast cancer and its relation to angiogenesis. Cancer Res 1997; 57: 963–969.

54.

Soffietti

Trevisan

Ruda

. Targeted therapy in brain metastasis. Curr Opin Oncol 2012; 24: 679–686.

Supplementary Material

Please find the following supplemental material available below.

For Open Access articles published under a Creative Commons License, all supplemental material carries the same license as the article it is associated with.

For non-Open Access articles published, all supplemental material carries a non-exclusive license, and permission requests for re-use of supplemental material or any part of supplemental material shall be sent directly to the copyright owner as specified in the copyright notice associated with the article.

0.00 MB

0.13 MB

0.15 MB

0.14 MB

0.32 MB