Purpose:

Invasive lobular carcinoma (ILC) represents the second most common histologic breast cancer subtype after invasive ductal carcinoma (IDC). While primary ILC has been extensively studied, metastatic ILC has been poorly characterized at the genomic and immune level.

Experimental Design:

We retrospectively assembled the multicentric EuroILC series of matched primary and metastatic samples from 94 patients with estrogen receptor (ER)-positive ILC. Stromal tumor-infiltrating lymphocytes (sTILs) were assessed by experienced pathologists. Targeted sequencing and low pass whole-genome sequencing were conducted to detect mutations and copy-number aberrations (CNAs). We compared the frequencies of the alterations in EuroILC with those from patients with ER-positive metastatic ILC (n = 135) and IDC (n = 563) from MSK-IMPACT.

Results:

Low sTIL levels were observed in ILC metastases, with higher levels in the mixed nonclassic histology. Considering ILC metastases from EuroILC and MSK-IMPACT, we observed that >50% of tumors harbor genomic alterations that have previously been associated with endocrine resistance. A matched primary/metastasis comparison in EuroILC revealed mutations (AKT1, ARID1A, ESR1, ERBB2, or NF1) and CNAs (PTEN or NF1 deletion, CYP19A1 amplification) associated with endocrine resistance that were private to the metastasis in 22% (7/32) and 19% (4/21) of patients, respectively. An increase in CDH1, ERBB2, FOXA1, and TBX3 mutations, in CDH1 deletions and a decrease in TP53 mutations was observed in ILC as compared with IDC metastases.

Conclusions:

ILC metastases harbor genomic alterations that may potentially explain endocrine resistance in a large proportion of patients, and present genomic differences as compared with IDC metastases.

Translational Relevance

Invasive lobular carcinoma (ILC) represents the second most common histologic breast cancer subtype after invasive ductal carcinoma (IDC) and mostly expresses the estrogen receptor (ER). Several genomic alterations have been associated with endocrine resistance in ER-positive breast cancer. However, so far, it was unknown whether these alterations are preexisting or acquired after endocrine treatment, and whether their frequencies differ between ILC and IDC metastases. Here, we present evidence of possibly acquired alterations associated with endocrine resistance during disease evolution in approximately one of five patients with ILC. In addition, the genomic landscape of metastatic ILC does present some clinically relevant differences with regard to metastatic IDC. While the present results still need to be confirmed in larger cohorts of patients with metastatic ILC, they support the need to characterize the metastatic disease, either using a metastatic biopsy or circulating tumor DNA, to refine treatment for these patients.

Invasive lobular carcinoma (ILC) accounts for up to 15% of all invasive breast cancers and represents the second most common histologic subtype of invasive breast cancer after invasive breast carcinoma of no special type, commonly referred to as invasive ductal carcinoma (IDC; ref. 1). Lobular tumors are typically characterized by small discohesive cells that lack membranous E-cadherin expression. These tumors generally express the estrogen receptor (ER) and lack HER2 amplification. Patients with ILC tend to relapse later (2) and present different metastatic pattern as compared with patients with IDC (3, 4).

Over the past few years, significant efforts have been done to characterize primary lobular tumors at the genomic, gene expression, and immune level (5–9). Genomic studies have provided a more precise estimation of the genomic alterations present in CDH1, the gene coding for E-cadherin (5–8): approximately 90% of ILC samples have heterozygous CDH1 deletion, mostly characterized by the loss of the long arm of chromosome 16, and approximately 60% exhibit a mutation in the other allele. Researchers have recently used this characteristic as a potential therapeutic opportunity (10). Indeed, inhibiting ROS1 is synthetically lethal in cells with CDH1 defects, and this strategy is currently being explored in a phase II trial (ClinicalTrials.gov Identifier: NCT03620643). In addition, the various genomic studies have further highlighted the increased prevalence of mutations in ER+/HER2 primary ILC in several genes such as AKT1, ARID1A, CDH1, FOXA1, HER2, HER3, PIK3CA, and TBX3 genes as compared with ER+/HER2 primary IDC (5–8). The gene expression studies have also identified different gene expression subtypes within primary lobular tumors, some with prognostic relevance (6, 7, 11). And finally, ER+ lobular tumors have been shown to have significantly lower levels of stromal tumor-infiltrating lymphocytes (sTILs) as compared with their ductal counterparts (9, 12).

Several large initiatives have recently genomically characterized metastases from patients with breast cancer. These include clinical trials, institutional programs, and patient-driven initiatives (13–16). While these were not focused on ILC, they have generated and released molecular data on a significant number of these tumors. So far, to the best of our knowledge, only two publications specifically reported on ILC metastases (17, 18). A recent study used data generated by Foundation Medicine on unmatched primary or metastatic samples from patients with metastatic ILC and reported a critical role of NF1 in mediating endocrine resistance (17). Our group investigated ESR1 mutations in metastatic samples from patients with ER+ ILC and reported a similar frequency and distribution as compared with metastatic samples from patients with ER+ IDC (18).

While immunotherapy based on checkpoint inhibition is currently being tested in patients with metastatic ILC (ClinicalTrials.gov Identifier: NCT03147040), the levels of immune infiltration in ILC metastases remain undetermined. Here, we therefore first centrally characterized the levels of sTILs in a unique European multicentric series of matched primary and metastatic samples from patients with metastatic ER+ ILC, further referred to as EuroILC. Second, we aimed at characterizing the samples from EuroILC at the genomic level to unravel endocrine resistance and disease progression. We further considered the publicly available genomic data from patients with metastatic ER-positive IDC and ILC from the MSK-IMPACT initiative (13) to identify differences in genomic landscapes of IDC and ILC metastases.

Patients with EuroILC and samples

The patients and samples of this cohort have already been described previously (18). In brief, we considered patients with metastatic ILC from six different European hospitals (Institut Jules Bordet-Brussels, Cliniques Universitaires Saint-Luc-Brussels, GZA Ziekenhuizen-Antwerp, Institut Paoli-Calmettes-Marseille, Institut Curie-Paris, Istituto Europeo di Oncologia-Milan) for which primary and metastatic samples were available. In addition, samples had to fulfil the following criteria: (i) no distinct invasive neoplastic components other than ILC at central revision (mixed ductal-lobular tumors were excluded); (ii) ER+ status of the primary tumor; (iii) minimal tumor cellularity of 20% (if <20%, then only considered if macrodissection could be done); and, for the DNA analyses (iv) availability of >100 ng of DNA from a formalin-fixed paraffin-embedded (FFPE) block of the primary tumor, metastasis, and noninvaded tissue. Priorities in terms of DNA characterization for this project were: first, ddPCR for detecting ESR1 mutations (18); second, targeted sequencing; and third, low pass whole-genome sequencing to detect copy-number changes and genomic instability. The project has been approved by the ethics committee of the Institut Jules Bordet (N°2504). Given the retrospective nature of the study, the ethics committee did not require the patients to sign an informed consent. The study was performed in accordance with the Declaration of Helsinki.

Histopathologic characterization

Central pathology review was assessed by D. Larsimont for histologic type and grade. ER, PgR, HER2, and Ki67 scores were retrieved from the pathology reports. sTILs were assessed by two experienced breast cancer pathologists (R. Salgado and G. van den Eynden) using the international guidelines (19). The average of the two scores was considered as the final value for that sample. sTIL values were log transformed and when several primary or metastatic samples were available from the same lesion, the median value was considered.

Targeted sequencing

A TruSeq Custom Amplicon panel of 20 genes frequently mutated in ILC and/or ER+ breast cancer in general was designed using DesignStudio from Illumina: AKT1, ARID1A, CDH1, ERBB2, ERBB3, ESR1, FOXA1, GATA3, IGF1R, JAK2, MAP2K4, MAP3K1, NF1, PIK3CA, PTEN, RB1, RUNX1, STAT3, TBX3, and, TP53. A total of 195 tumor samples from 73 patients were sequenced using the Amplicon DS technology from Illumina together with a matched normal sample. Both DNA strands were processed independently, considering the intersection of the called variants between the two strands to limit the noise generated by the FFPE artifacts. Reads were aligned on hg19 using bwa (20). Samples achieving a median coverage of 150× were kept for downstream analyzes. Normal samples were merged per patient when possible to increase coverage, all normal samples were merged in a pool-of-normal sample to serve as a germline reference for patients without good quality matched normal sample available (15/73 = 20% patients). Manta (21) followed by Strelka2 (22) were used for the calling of both single base substitutions and small somatic insertions–deletions. Possible driver mutations were identified according to Desmedt and colleagues, 2016 (5).

Low pass whole-genome sequencing and analysis of copy-number data

A total of 165 tumor samples from 63 patients were sequenced with Illumina to an average target coverage of 0.5×. Samples were cosegmented per patient using Copynumber (23), and copy-number gains and losses were assessed using ABSOLUTE (24). Solutions were manually selected. In case of inconsistency between the matched primary and metastatic samples, the most reliable sample according to its purity was used to fit the other samples. Samples having a purity of at least 30% were kept for downstream analyses (Supplementary Methods). For the comparisons, we focused on the genes that were reported to be potential “drivers” by Nik-Zainal and colleagues (25), as well as CYP19A amplification (26). All the aberrations private to the either one of the matched primary or metastatic samples were manually reviewed. Chromosomal instability (CIN) scores were computed as the ratio between the number of base pair either amplified or deleted over the total number of base pairs analyzed in the cohort.

Publicly available data

MSK-IMPACT data (13), including clinical data, mutations, and raw segmentation data were retrieved from cBioPortal in November 2018 (27). Only female ER-positive patients having experienced a relapse event were selected. Patients with ILC and IDC were identified according to their oncotree codes (ILC and IDC, BRCANOS, BRCNOS, CSNOS, respectively). Possible driver mutations and copy-number profiles were determined using the same protocols as for EuroILC.

Statistical analysis

Association with clinicopathologic variables, mutations and copy-number alteration frequencies were assessed using Fisher exact tests. Association between continuous sTILs and categorical variables was assessed with Wilcoxon and Kruskal–Wallis tests when having two or more than two categories, respectively. When appropriate, a paired Wilcoxon test was performed to consider the matched status of the primary and metastatic samples. Agreement between the two pathologists in scoring the sTILs was assessed using the concordance correlation coefficient (CCC) and its confidence interval using the DescTools R package. Association between genomic alterations and clinicopathologic variables were assessed using logistic regression. Median follow-up was computed using the reverse Kaplan–Meier estimator. All tests were considered significant for P value < 0.05. When appropriate, multiple testing correction were applied using FDR and q values <0.05 were considered significant. All analyses were performed under R 3.5.2.

Availability of data and materials

The BAM files generated in this study are available on Genome-Phenome Archive (EGA) platform (https://ega-archive.org) under accession number EGAS00001004641. In addition, processed data including clinical, mutation, and copy-number aberration (CNA) data used in this analysis are available in Supplementary Tables S1, S3, and S4, respectively. More detailed methods can be found in the Data Supplement.

Patient and sample characteristics

For the EuroILC cohort, we retrospectively selected 125 patients with metastatic ILC for which both primary and metastatic samples were available from six European institutions. After central pathology review, 94 patients were eligible for matched primary/metastasis sTIL assessment. Sufficient DNA was available for 80 patients to also assess ESR1 mutations using droplet digital PCR (18). In addition, in 73 and 63 patients, respectively, there was sufficient DNA to perform targeted and low pass whole-genome sequencing (Supplementary Fig. S1). The clinical and pathologic characteristics of the EuroILC cohort (n = 94) are presented in Table 1 and Supplementary Table S1. In summary, considering the patients eligible for sTIL assessment, 56% of the patients were menopaused at primary diagnosis, 76% had a tumor >2 cm, 69% had axillary lymph node involvement, 44% of the tumors were of the nonclassical histologic subtype, 19% were high grade, 85% expressed the progesterone receptor (PgR), and 8% were HER2-amplified. The grading distribution according to the histologic subtypes was expected and as follows: 31%, 63%, and 6% of the patients having classic ILC of grade 1, 2, 3, respectively; and 3%, 62%, and 35% of the patients having nonclassical ILC of grade 1, 2, and 3, respectively. Eighty-four percent and 46% of the patients received adjuvant endocrine therapy and adjuvant chemotherapy, respectively. Loss of hormonal receptor status was observed in the metastatic samples in 11% and 38% of the patients for ER and PgR, respectively. The vast majority (84%) of the metastatic biopsies were taken within the year after metastasis diagnosis.

Table 1.

Patient and sample characteristics of the EuroILC, MSK-IMPACT ILC, and IDC cohorts.

EuroILC (94 pts)MSK-IMPACT ILC (135 pts)MSK-IMPACT IDC (563 pts)EuroILC vs. MSK ILCEuroILC vs. MSK IDCMSK ILC vs. IDC
Age at diagnosis 
 <50 28 (30%) 51 (38%) 305 (54%) 0.26 <0.001 <0.001 
 ≥50 66 (70%) 84 (62%) 258 (46%)    
Menopausal status 
 Pre/peri 41 (44%) 67 (50%) 351 (63%) 0.43 <0.001 <0.01 
 Post 53 (56%) 67 (50%) 209 (37%)    
 Missing    
Histologic primary tumor size 
 <2 cm 21 (24%) 43 (35%) 210 (44%) 0.04 <0.001 0.08 
 ≥2 cm 67 (76%) 79 (65%) 263 (56%)    
 Missing 13 90    
Nodal status 
 Negative 29 (31%) 40 (33%) 194 (40%) 0.88 0.13 0.18 
 Positive 64 (69%) 82 (67%) 293 (60%)    
 Missing 13 76    
Histologic subtype 
 Classic 51 (56%) NA NA NA  NA 
 Nonclassic 40 (44%) NA NA    
 Missing      
Histologic grade (primary) 
 G1/G2 75 (81%) 34 (47%) 139 (27%) <0.001 <0.001 <0.001 
 G3 18 (19%) 38 (53%) 382 (73%)    
 Missing 63 42    
PgR status (primary) 
 Negative 14 (15%) 23 (18%) 109 (19%) 0.72 0.47 0.71 
 Positive 77 (85%) 108 (82%) 453 (81%)    
 Missing    
HER2 status (primary) 
 Negative 78 (92%) 119 (94%) 448 (86%) 0.57 0.22 0.01 
 Positive 7 (8%) 7 (6%) 70 (14%)    
 Missing 45    
Adjuvant chemotherapy 
 No 51 (54%) 77 (57%) 292 (52%) 0.69 0.74 0.29 
 Yes 43 (46%) 58 (43%) 271 (48%)    
 Missing    
Adjuvant endocrine therapy 
 No 15 (16%) 31 (23%) 182 (32%) 0.24 0.001 0.04 
 Yes 78 (84%) 104 (77%) 381 (68%)    
 Missing    
Adjuvant radiotherapy 
 No 27 (30%) NA NA NA NA NA 
 Yes 64 (70%) NA NA    
 Missing      
Time to relapse 
 Median 4.69 years 4.65 years 3.53 years    
De novo metastatic 14 (15%) 24 (18%) 109 (19%) 0.58 0.17 0.24 
 Relapsed <5 years 34 (36%) 46 (34%) 235 (42%)    
 Relapsed >5 but <10 years 31 (33%) 36 (27%) 126 (22%)    
 Relapsed >10 years 15 (16%) 29 (21%) 93 (17%)    
ER status (metastasis) 
 Negative 9 (11%) 16 (13%) 71 (13%) 0.83 0.72 1.00 
 Positive 73 (89%) 108 (87%) 471 (87%)    
 Missing 12 11 21    
PgR status (metastasis) 
 Negative 38 (48%) 59 (50%) 249 (47%) 0.89 0.81 0.61 
 Positive 41 (52%) 60 (50%) 285 (53%)    
 Missing 15 16 29    
HER2 status (metastasis) 
 Negative 75 (95%) 109 (92%) 463 (88%) 0.41 0.08 0.42 
 Positive 4 (5%) 10 (8%) 61 (12%)    
 Missing 15 16 39    
Timing metastatic biopsy 
 <1 year diagnosis 75 (84%) NA NA NA NA NA 
 Later 14 (16%) NA NA    
 Missing      
ET before metastatic sampling 
 SERM 39 (41%) 34 (25%) 185 (33%) 0.01 <0.001 <0.001 
 AI only 20 (21%) 39 (29%) 110 (20%)    
 SERM and AI 28 (30%) 35 (26%) 92 (16%)    
 No ET 7 (7%) 27 (20%) 176 (31%)    
Duration ET before metastatic sampling 
 <2 years 17 (20%) 16 (15%) 101 (26%) 0.54 0.36 0.01 
 2–4 years 20 (23%) 22 (20%) 94 (24%)    
 >4 years 50 (57%) 70 (65%) 192 (50%)    
Deceased 
 No 43 (46%) 99 (73%) 424 (75%) <0.001 <0.001 0.66 
 Yes 51 (54%) 36 (27%) 139 (25%)    
EuroILC (94 pts)MSK-IMPACT ILC (135 pts)MSK-IMPACT IDC (563 pts)EuroILC vs. MSK ILCEuroILC vs. MSK IDCMSK ILC vs. IDC
Age at diagnosis 
 <50 28 (30%) 51 (38%) 305 (54%) 0.26 <0.001 <0.001 
 ≥50 66 (70%) 84 (62%) 258 (46%)    
Menopausal status 
 Pre/peri 41 (44%) 67 (50%) 351 (63%) 0.43 <0.001 <0.01 
 Post 53 (56%) 67 (50%) 209 (37%)    
 Missing    
Histologic primary tumor size 
 <2 cm 21 (24%) 43 (35%) 210 (44%) 0.04 <0.001 0.08 
 ≥2 cm 67 (76%) 79 (65%) 263 (56%)    
 Missing 13 90    
Nodal status 
 Negative 29 (31%) 40 (33%) 194 (40%) 0.88 0.13 0.18 
 Positive 64 (69%) 82 (67%) 293 (60%)    
 Missing 13 76    
Histologic subtype 
 Classic 51 (56%) NA NA NA  NA 
 Nonclassic 40 (44%) NA NA    
 Missing      
Histologic grade (primary) 
 G1/G2 75 (81%) 34 (47%) 139 (27%) <0.001 <0.001 <0.001 
 G3 18 (19%) 38 (53%) 382 (73%)    
 Missing 63 42    
PgR status (primary) 
 Negative 14 (15%) 23 (18%) 109 (19%) 0.72 0.47 0.71 
 Positive 77 (85%) 108 (82%) 453 (81%)    
 Missing    
HER2 status (primary) 
 Negative 78 (92%) 119 (94%) 448 (86%) 0.57 0.22 0.01 
 Positive 7 (8%) 7 (6%) 70 (14%)    
 Missing 45    
Adjuvant chemotherapy 
 No 51 (54%) 77 (57%) 292 (52%) 0.69 0.74 0.29 
 Yes 43 (46%) 58 (43%) 271 (48%)    
 Missing    
Adjuvant endocrine therapy 
 No 15 (16%) 31 (23%) 182 (32%) 0.24 0.001 0.04 
 Yes 78 (84%) 104 (77%) 381 (68%)    
 Missing    
Adjuvant radiotherapy 
 No 27 (30%) NA NA NA NA NA 
 Yes 64 (70%) NA NA    
 Missing      
Time to relapse 
 Median 4.69 years 4.65 years 3.53 years    
De novo metastatic 14 (15%) 24 (18%) 109 (19%) 0.58 0.17 0.24 
 Relapsed <5 years 34 (36%) 46 (34%) 235 (42%)    
 Relapsed >5 but <10 years 31 (33%) 36 (27%) 126 (22%)    
 Relapsed >10 years 15 (16%) 29 (21%) 93 (17%)    
ER status (metastasis) 
 Negative 9 (11%) 16 (13%) 71 (13%) 0.83 0.72 1.00 
 Positive 73 (89%) 108 (87%) 471 (87%)    
 Missing 12 11 21    
PgR status (metastasis) 
 Negative 38 (48%) 59 (50%) 249 (47%) 0.89 0.81 0.61 
 Positive 41 (52%) 60 (50%) 285 (53%)    
 Missing 15 16 29    
HER2 status (metastasis) 
 Negative 75 (95%) 109 (92%) 463 (88%) 0.41 0.08 0.42 
 Positive 4 (5%) 10 (8%) 61 (12%)    
 Missing 15 16 39    
Timing metastatic biopsy 
 <1 year diagnosis 75 (84%) NA NA NA NA NA 
 Later 14 (16%) NA NA    
 Missing      
ET before metastatic sampling 
 SERM 39 (41%) 34 (25%) 185 (33%) 0.01 <0.001 <0.001 
 AI only 20 (21%) 39 (29%) 110 (20%)    
 SERM and AI 28 (30%) 35 (26%) 92 (16%)    
 No ET 7 (7%) 27 (20%) 176 (31%)    
Duration ET before metastatic sampling 
 <2 years 17 (20%) 16 (15%) 101 (26%) 0.54 0.36 0.01 
 2–4 years 20 (23%) 22 (20%) 94 (24%)    
 >4 years 50 (57%) 70 (65%) 192 (50%)    
Deceased 
 No 43 (46%) 99 (73%) 424 (75%) <0.001 <0.001 0.66 
 Yes 51 (54%) 36 (27%) 139 (25%)    

Note: Numbers represent the counts and percentage are shown in brackets. Differences in distributions between cohorts are screened by Fisher tests and P values are reported in the right columns. Of note, when multiple metastatic samples were available for a patient, we arbitrarily reported the ER, PgR, and HER2 status from the latest one.

Abbreviations: AI, aromatase inhibitors; ER, estrogen receptor; ET, endocrine therapy; G, grade, PgR, progesterone status; pts, patients; SERM, specific estrogen-receptor modulators.

sTILs in ILC metastases

To estimate the immune infiltration in the primary and metastatic samples from the EuroILC cohort, two experienced pathologists independently scored sTILs using acknowledged guidelines (19). A good concordance was observed (CCC: 0.89, 95% CI: 0.86–0.91), leading us to consider the average of the two scores as the final value. We first compared the distribution of sTILs in the primary tumors from EuroILC with the primary tumors from the consecutive multicentric series that were previously analyzed by the same pathologists (9). We observed statistically lower levels in the primary tumor samples from EuroILC [unpaired Wilcoxon test, P < 0.001, Fig. 1A, medians (interquartile range (IQR)) of 5% (3–7) and 1% (1–7) for Desmedt and colleagues (9) and EuroILC's primaries, respectively]. A paired analysis revealed no significant difference in sTIL levels between the paired primary and metastatic samples from EuroILC [paired Wilcoxon test, P = 0.11, Fig. 1A, median (IQR) of 1% (1–5) for EuroILC's metastases]. Higher sTILs in the primary tumor samples from EuroILC were associated with younger age at diagnosis, and with mixed nonclassic (which include the so-called pleomorphic tumors, and should not be confused with the mixed ductal-lobular tumors) and trabecular histology (Supplementary Fig. S2A). Higher sTILs in the metastatic samples showed a trend toward association with the mixed nonclassic histology (Supplementary Fig. S2B). sTIL infiltration in the ILC metastases did not differ according to the metastatic site (Fig. 1B). Immune infiltration is illustrated in Fig. 1C and D for a metastasis in the liver and serosal membrane, respectively.

Figure 1.

sTIL landscape in ILC specimens. A, Comparison of EuroILC primary sTILs (58 patients) with the Desmedt and colleagues (2018) ILC primary sTILs (577 patients; ref. 9) and EuroILC metastasis sTILs (67 patients). Significance of the differences are assessed by a Wilcoxon unpaired test between EuroILC primaries and Desmedt and colleagues 2018 ILC primaries and a paired test between EuroILC primaries and EuroILC metastases. The median value is considered if several samples are available per patient. B, sTIL infiltration according to the metastatic site. Scanned hematoxylin and eosin slides from two EuroILC metastases: liver sample (4M2) with 70% sTILs (C); serosal membrane sample (24M1) with 22.5% sTILs (D). GI, gastrointestinal; mb, membrane.

Figure 1.

sTIL landscape in ILC specimens. A, Comparison of EuroILC primary sTILs (58 patients) with the Desmedt and colleagues (2018) ILC primary sTILs (577 patients; ref. 9) and EuroILC metastasis sTILs (67 patients). Significance of the differences are assessed by a Wilcoxon unpaired test between EuroILC primaries and Desmedt and colleagues 2018 ILC primaries and a paired test between EuroILC primaries and EuroILC metastases. The median value is considered if several samples are available per patient. B, sTIL infiltration according to the metastatic site. Scanned hematoxylin and eosin slides from two EuroILC metastases: liver sample (4M2) with 70% sTILs (C); serosal membrane sample (24M1) with 22.5% sTILs (D). GI, gastrointestinal; mb, membrane.

Close modal

Clinicopathologic comparison of patient and sample characteristics from EuroILC and MSK-IMPACT

We compared the clinical and pathologic characteristics between patients with metastatic ER+ ILC from EuroILC and MSK-IMPACT (Table 1; Supplementary Fig. S3; Supplementary Table S2). The series were comparable with some exceptions: first, more than half of the patients with MSK-IMPACT (53%) have high-grade tumors as compared with only 19% of the patients with EuroILC, and, second, consistent with the retrospective and prospective setting of the two cohorts, more than half of the patients from EuroILC (54%) already died as compared with 28% in MSK-IMPACT. This is consistent with the difference in median follow-up observed for overall survival between the cohorts: 12.2 years for EuroILC, 5.1 and 8.1 years for the patients with IDC and ILC metastases from MSK-IMPACT, respectively. We also compared the sites of metastatic samples from EuroILC with the metastatic samples from patients with ILC from MSK-IMPACT (Supplementary Fig. S4). The distribution was very similar with the exception of samples from serosal membranes metastases that were more represented in EuroILC and biopsies from chest wall metastases that were more prevalent in MSK-IMPACT. We next compared the sites of metastatic samples between patients with ILC (EuroILC and MSK-IMPACT) and IDC (MSK-IMPACT), focusing only on patients with ER+ tumors, and consistently observed for ILC more samples from the bone, reproductive organs, and gastrointestinal tracts, and less from the liver, lung, distant lymph nodes, and soft tissues (Supplementary Fig. S4).

Genomic comparison of matched primary and metastatic EuroILC samples

The individual clinical and pathologic characteristics of the 43 patients with EuroILC that have genomic data and passed quality control for both their primary and metastatic samples are represented in Fig. 2. For >50% of patients, either the primary or the metastatic sample(s) did not pass QC because the DNA was too degraded (Supplementary Fig. S1). We first compared the mutational landscape between matched primary and metastatic samples (32 patients). Of note, when multiple primary samples per patient were sequenced, which was the case in 16/32 (50%) of patients, then the mutation was considered to be present if it was detected in at least one of the samples. A similar rule was applied for the metastatic samples, concerning also 16/32 (50%) patients, with multiple metastatic samples being mostly available from the same metastatic lesion. As displayed in the heatmap in Fig. 3A, genes mutated in >10% of either primary and/or metastatic samples were in decreasing order: CDH1, PIK3CA, TBX3, MAP3K1, TP53, GATA3, ESR1, ERBB2, ARID1A, IGF1R, FOXA1, and AKT1. Lolipop plots representing the distribution of mutations in some of these genes are illustrated in Fig. 3B. We observed mutations private to the metastatic sample(s), that is, mutations only detected in the metastatic but not in the matched primary sample(s), in CDH1 in 7/32 (with 2/7 that also have another shared CDH1 mutation in their primary and metastatic samples), in ARID1A, ERBB2, and ESR1 in 2/32 (6%) of the patients, and in AKT1, GATA3, MAPK3K1, NF1, RB1, and RUNX1 in 1/32 (3%) of the patients. Mutations private to the metastasis were also seen in two patients where the metastatic sample lost ER expression (patient 32 and 36). Of note, loss of several mutations that were present in the primary tumor was observed in the metastatic samples from some patients. This could be explained by tumor heterogeneity, since some of these mutations were only present at a low allele frequency (such as ERBB3 D501N mutation in patient 9, seen at a VAF of 8% only in one of the two primary samples, not in the metastatic sample) or in only some of the primary tumor samples (ERBB2 L755S mutation in patient 1, seen at a VAF of 51% in only one of the five primary samples, not in the metastatic sample) (Supplementary Fig. S5). With regard to copy-number alterations (Fig. 3C and D), results were available for 21 matched primary metastasis patients. Here again, as for the mutations, we observed copy-number alterations private to the metastatic sample(s), such as MAP2K4 and NCOR1 deletions in four of 21 (19%) patients, TP53 deletion in three of 21 (14%) patients, PTEN and AKT1 deletions in two of 21 (10%) patients, as well as CCND1 and CCNE1 amplifications in two of 21 (10%) patients (Supplementary Fig. S6). Figure 4 represents the disease evolution and the treatments received of three patients, together with the private alterations observed in each sample.

Figure 2.

Overview of the patient and sample characteristics from EuroILC for the patients with genomic data. The bar plots show the clinical course of the 43 patients with targeted sequencing and/or CNA data that passed QC for both their primary and metastatic samples, where stars indicate death. In the heatmap, green represents positive or high values (age >= 50, size >= 2 cm, positive node, grade III, ER, PgR, HER2+, patient has a nonclassic subtype, patient is de novo metastatic, patient has the treatment type, patient was processed for targeted sequencing, and copy-number analysis), while gray represents the contrary and white the missing values. adj, adjuvant; chemo, chemotherapy; ER, estrogen receptor; HT, hormone therapy; met, metastatic; PgR, progesterone receptor; prim, primary; QC, quality control; RT, radiotherapy; CNA, copy-number aberration; sTILs, stromal tumor-infiltrating lymphocytes.

Figure 2.

Overview of the patient and sample characteristics from EuroILC for the patients with genomic data. The bar plots show the clinical course of the 43 patients with targeted sequencing and/or CNA data that passed QC for both their primary and metastatic samples, where stars indicate death. In the heatmap, green represents positive or high values (age >= 50, size >= 2 cm, positive node, grade III, ER, PgR, HER2+, patient has a nonclassic subtype, patient is de novo metastatic, patient has the treatment type, patient was processed for targeted sequencing, and copy-number analysis), while gray represents the contrary and white the missing values. adj, adjuvant; chemo, chemotherapy; ER, estrogen receptor; HT, hormone therapy; met, metastatic; PgR, progesterone receptor; prim, primary; QC, quality control; RT, radiotherapy; CNA, copy-number aberration; sTILs, stromal tumor-infiltrating lymphocytes.

Close modal
Figure 3.

Genomic landscape of the primary/metastatic matched samples from the EuroILC cohort. A, C, D, Green, red, and blue in the heatmap represent alterations private to primary tumor, at least one private to the metastasis, and present in both the primary and the metastasis, respectively. Colors of the patient numbers refer to the histologic type of the primary tumor of the patient with orange, turquoise, brown, purple referring to classic, mixed nonclassic, solid, and trabecular histologic types, respectively. A, Mutational landscape from targeted sequencing approach concerning 32 patients. B, Lollipop representations of the most altered genes in EuroILC cohort. Green, black, and brown colors are for missense, truncating, in-frame mutations, respectively. The y-axis is the alteration frequency among patients. C and D, Deletions and amplification, respectively, observed from the low pass whole-genome sequencing on 21 patients. Colors of the lines (E) and dots (F) refer to the histologic type of the primary breast cancer tumor of the patient with same color code as above. E, CIN score evolution from primary to matched metastatic sample in EuroILC. Median scores are taken when several samples are available per patient, increases and decreases are depicted in solid and dashed lines, respectively. F, Differences in CIN scores between metastases and primaries against elapsing time between sampling of the primary and the metastasis. Black line depicts the linear regression between the two variables. G, CIN scores distributions between primary and metastatic samples in green and red, respectively, across the different studied cohorts. There are 43, 41, 33, 94, 174, and 361 patients available for EuroILC primaries, EuroILC metastases, MSK-IMPACT ILC primaries, MSK-IMPACT ILC metastases, MSK-IMPACT IDC primaries, and MSK-IMPACT IDC metastases regarding copy-number analyses, respectively. CIN, chromosomal instability; M, metastatic sample; P, primary sample.

Figure 3.

Genomic landscape of the primary/metastatic matched samples from the EuroILC cohort. A, C, D, Green, red, and blue in the heatmap represent alterations private to primary tumor, at least one private to the metastasis, and present in both the primary and the metastasis, respectively. Colors of the patient numbers refer to the histologic type of the primary tumor of the patient with orange, turquoise, brown, purple referring to classic, mixed nonclassic, solid, and trabecular histologic types, respectively. A, Mutational landscape from targeted sequencing approach concerning 32 patients. B, Lollipop representations of the most altered genes in EuroILC cohort. Green, black, and brown colors are for missense, truncating, in-frame mutations, respectively. The y-axis is the alteration frequency among patients. C and D, Deletions and amplification, respectively, observed from the low pass whole-genome sequencing on 21 patients. Colors of the lines (E) and dots (F) refer to the histologic type of the primary breast cancer tumor of the patient with same color code as above. E, CIN score evolution from primary to matched metastatic sample in EuroILC. Median scores are taken when several samples are available per patient, increases and decreases are depicted in solid and dashed lines, respectively. F, Differences in CIN scores between metastases and primaries against elapsing time between sampling of the primary and the metastasis. Black line depicts the linear regression between the two variables. G, CIN scores distributions between primary and metastatic samples in green and red, respectively, across the different studied cohorts. There are 43, 41, 33, 94, 174, and 361 patients available for EuroILC primaries, EuroILC metastases, MSK-IMPACT ILC primaries, MSK-IMPACT ILC metastases, MSK-IMPACT IDC primaries, and MSK-IMPACT IDC metastases regarding copy-number analyses, respectively. CIN, chromosomal instability; M, metastatic sample; P, primary sample.

Close modal
Figure 4.

A–C, Clinical course of three selected patients with EuroILC. Aromatase inhibitors, ESR inhibitors, chemotherapy, and radiotherapy are represented in dark green, green, red, and dark red, respectively. Private alterations including mutations, short indels, copy-number gains, and losses are indicated under the corresponding processed samples. Samples available for sequencing analysis are shown in blue boxes. M, metastatic sample; Met, metastasis; P, primary sample.

Figure 4.

A–C, Clinical course of three selected patients with EuroILC. Aromatase inhibitors, ESR inhibitors, chemotherapy, and radiotherapy are represented in dark green, green, red, and dark red, respectively. Private alterations including mutations, short indels, copy-number gains, and losses are indicated under the corresponding processed samples. Samples available for sequencing analysis are shown in blue boxes. M, metastatic sample; Met, metastasis; P, primary sample.

Close modal

We further computed chromosomal instability for all primary and metastatic samples, and computed for each patient the change in instability (Fig. 3EG). We observed a positive but not statistically significant association between the primary/metastatic time difference in sampling and the primary/metastatic difference in chromosomal instability (Spearman rho = 0.34, P = 0.14), suggesting that the longer the disease evolution lasts, the more the metastasis differs from the primary tumor.

Comparison of the genomic landscape of ILC and IDC metastases from EuroILC and MSK-IMPACT

We started by comparing the genomic alterations present in the primary and metastatic ILC samples from EuroILC (from 54 and 40 patients, respectively) and MSK-IMPACT (from 51 and 135 patients, respectively), to see whether these were comparable despite the clinical and pathologic differences reported above (Fig. 5A and B; Supplementary Fig. S7). Overall, ILC samples from both cohorts were comparable. Nevertheless, we observed an increased frequency in copy-number losses in AKT1, BRCA2, and RB1 in the primary samples from EuroILC as compared with the primary ILC samples from MSK-IMPACT, and an increased frequency in IGF1R mutations and copy-number losses in AKT1 and MAP3K1 in the metastases from EuroILC as compared with those from MSK-IMPACT. To identify genes that are differentially altered between ER+ ILC and IDC metastases, we first compared the mutations and CNAs present in the samples from the patients with EuroILC with the ones from the MSK-IMPACT patients with IDC (260 with primary samples and 563 with metastatic samples; Fig. 5C and D; Supplementary Fig. S7). We then repeated the comparison between the ILC and IDC metastases from MSK-IMPACT (Fig. 5E and F; Supplementary Fig. S7). In terms of mutations, the two comparisons consistently identified a significant increased prevalence of mutations affecting CDH1, ERBB2, FOXA1, and TBX3, and a decreased prevalence in TP53 mutations in ILC metastases as compared with IDC metastases. In addition, GATA3 mutations were further more prevalent in IDC metastases, although it only reached statistical significance in the MSK-IMPACT comparison. Finally, IGF1R mutations were significantly more prevalent in ILC metastases but only when considering the EuroILC cohort. No difference in prevalence was, however, observed with regard to PIK3CA or ESR1 mutations. With regard to copy-number aberrations, CDH1 deletions were, as expected, consistently more frequent, in the primary and metastatic samples from ILC as compared with IDC. Additional copy-number differences were further observed but only in the comparison involving either the EuroILC or MSK-IMPACT ILC cohort (Supplementary Fig. S7). Of note, no difference was observed with regard to the prevalence of TP53 deletions.

Figure 5.

Mutational landscape of ILC as compared with IDC in primaries and metastatic samples. A–F, Percentage of patients having the alterations are shown on the axes. Level of significance is color coded according to the P value adjusted for multiple testing using FDR, gray, pink, red for q-value < 0.01, <0.1, >= 0.1, respectively. Genes differentially altered at least once between two series are labeled in all comparisons. ESR1 and PIK3CA were not differentially altered but are labeled for information. There are 54, 40, 51, 135, 260, and 563 patients available for EuroILC primaries, EuroILC metastases, MSK-IMPACT ILC primaries, MSK-IMPACT ILC metastases, MSK-IMPACT IDC primaries, and MSK-IMPACT IDC metastases regarding mutation analyses, respectively. FDR, false discovery rate; IDC, invasive ductal carcinoma; ILC, invasive lobular breast cancer; M, metastatic; P, primary.

Figure 5.

Mutational landscape of ILC as compared with IDC in primaries and metastatic samples. A–F, Percentage of patients having the alterations are shown on the axes. Level of significance is color coded according to the P value adjusted for multiple testing using FDR, gray, pink, red for q-value < 0.01, <0.1, >= 0.1, respectively. Genes differentially altered at least once between two series are labeled in all comparisons. ESR1 and PIK3CA were not differentially altered but are labeled for information. There are 54, 40, 51, 135, 260, and 563 patients available for EuroILC primaries, EuroILC metastases, MSK-IMPACT ILC primaries, MSK-IMPACT ILC metastases, MSK-IMPACT IDC primaries, and MSK-IMPACT IDC metastases regarding mutation analyses, respectively. FDR, false discovery rate; IDC, invasive ductal carcinoma; ILC, invasive lobular breast cancer; M, metastatic; P, primary.

Close modal

In this study, we assembled retrospectively matched primary and metastatic samples from patients with metastatic ILC to evaluate the levels of immune infiltrates in the metastatic disease and to investigate de novo and acquired genomic alterations associated with endocrine resistance. We further aimed at comparing the genomic landscape of metastases from patients with ER+/HER2 ILC and ER+/HER2 IDC, using the publicly available MSK-IMPACT series (13).

In EuroILC, we observed a loss of ER and PgR in the metastases from 11% and 38% of the patients, respectively. This is in line with what has been reported so far in the literature for endocrine-resistant breast cancer (28). With regard to the site of metastatic sampling in patients with ILC, our results are consistent with Sokol and colleagues (17) and in agreement with previous reports that demonstrated that after accounting for hormone receptor status, patients with ILC had less lung and liver metastases but more ovarian and gastrointestinal metastases both at first site and overall, as compared with patients with IDC (3, 4).

We previously characterized immune infiltrates in primary tumors from patients with ILC and reported lower levels of sTILs in ER+/HER2 ILC as compared with ER+/HER2 IDC (9, 12). Here, we observed even lower levels of sTILs in the primary tumors from the EuroILC cohort as compared with our previous consecutive series of primary ILC. There was no difference in sTIL levels between the primary and matched metastasi(e)s, all being globally very low. We nevertheless observed higher sTIL levels both in the primary tumors and the metastases of the mixed nonclassic histologic subtype, which encompasses the so-called pleomorphic tumors. There was no apparent distinction between the various metastatic sites, although numbers were too small to apply statistical tests. A more detailed characterization of these immune cells is needed to better identify patients who may benefit from immunotherapy.

Several somatic genomic alterations associated with endocrine resistance have been uncovered and have recently been summarized by Hanker and colleagues (29). In EuroILC, 21/40 (53%) of the patients were harboring at least one mutation in their metastases which was previously reported to be associated with endocrine resistance, such as AKT1 (5, 30), ARID1A (13, 31), ERBB2 (5, 13, 32), ESR1 (33, 34), FOXA1 (13), NF1 (13, 17), or PTEN (35) mutations. A similar frequency was observed in the MSK-IMPACT ILC cohort, namely, 56/135 (41%). Of interest, 7/32 (22%) of the patients from the EuroILC cohort had private point mutations in the metastatic sample(s) in the AKT1 (5, 30), ARID1A (13, 31), ERBB2 (5, 13, 32), ESR1 (33, 34), or NF1 (13, 17, 36) gene. If we consider CNAs, then 12/21 had at least one event associated with endocrine resistance, such as ERBB2, CYP19A1 (26), EGFR (13), FGFR1 (37), or MYC (13) amplification, or PTEN (35) or NF1 deletion. We further observed aberrations private to the metastasis in 4/21 (19%) patients, such as PTEN (35) or NF1 (13, 17, 36) deletion or CYP19A1 (26) amplification. While the fact of observing mutations only in metastases and not in the matched primary samples is suggestive of acquired resistance, this cannot be formally proved as minor subclones might have been present in a part of the primary tumor that has not been sequenced. Altogether, as several of these alterations are targetable (29), it implies that a characterization of the metastatic disease, either by sequencing a metastatic biopsy or circulating tumor DNA (ctDNA) as in the PlasmaMATCH trial (30), is necessary to personalize the treatment of patients with metastatic ILC.

Some observations could be made when comparing the genomic landscape of metastases from patients with ER+ ILC and IDC such as an enrichment of mutations affecting CDH1, ERBB2, FOXA1, and TBX3 genes, an increase in CDH1 deletions, and a decreased prevalence in TP53 mutations, but not TP53 deletions, in ILC metastases as compared with IDC metastases. Of interest, we observed an increased frequency of IGFR1 mutations in metastases from the EuroILC cohort. While these would still need to be functionally characterized, Nagle and colleagues demonstrated that the hyperactivation of the IGF1R pathway was more frequent in ILC as compared with IDC primary tumors and that this could result in increased sensitivity to IGF1R/InsR targeted therapy (38). We did not detect an increased prevalence of NF1 mutations or deletions in metastases from patients with ILC, as suggested by Sokol and colleagues (17). Furthermore, it is important to report that we did not observe a difference in the frequency of PIK3CA mutations between ILC and IDC metastases, a finding that is clinically relevant because the first PI3K α-specific inhibitor alpelisib is now available to treat patients with hormone receptor–positive/HER2 advanced breast cancer (39). In this context, it is however important to remember that previous studies have demonstrated that the PI3K/Akt pathway can be activated in ILC independently of oncogenic mutations in this pathway (6, 40).

Our study has several limitations. First, the immune characterization has here so far been limited to the standard evaluation of sTILs (19). A more detailed and spatial evaluation of each cell type from the tumor microenvironment is however needed (29). Second, for most of the patients only one metastatic lesion was biopsied. Acknowledging the intrapatient intermetastases genomic heterogeneity which has been highlighted in several autopsy-based studies (13, 41–47), we can reasonably assume to have missed some genomic alterations present in other metastatic sites or in other regions of the biopsied metastasis. Importantly, this intrapatient intermetastases heterogeneity has also been shown to affect genes involved in treatment resistance such as ESR1 and PTEN mutations (13, 41, 46). Third, the rather limited panel size for mutational analysis may have prevented us from discovering and/or confirming recent ILC-specific driver genes that are present at a low frequency in the tumors of patients with ILC, such as mutations in FGFR4 (48) or BRAF, CTCF, KRAS, and MAP2K1 (13). Fourth, our study did not consider additional nongenomic elements that could be associated with endocrine resistance in ILC, such as lipid metabolism which has previously been suggested as a potential mechanism by cell line experiments (49). Our results suggest however that somatic alterations could already explain endocrine resistance in more than half of the patients with metastatic ILC. A final limitation relates to the number of patients in EuroILC. This exemplifies the difficulty in assembling retrospectively a cohort of patients with ILC with sufficient and good-quality DNA from matched primary and metastatic samples. This represents, however, a general need and challenge for the general breast cancer community and not only for patients with ILC (29).

To conclude, first, while most of the ILC metastases present low or no immune infiltration, higher levels can be observed in tumors from the mixed nonclassic histology. Whether or not these patients represent potential candidates for immunotherapy requires additional investigation. Second, we do have evidence of possibly acquired alterations associated with endocrine resistance during disease evolution in approximately one of five patients with ILC. While the present results still need to be confirmed in larger cohorts of patients with metastatic ILC, they support the need to characterize the metastatic disease, either using a metastatic biopsy or ctDNA, to refine treatment for these patients. In addition, the identified genomic alterations may not only be associated with endocrine resistance, but could also predict resistance to additional treatments. For instance, PTEN alterations have recently been associated with resistance to PI3Kα and CDK4/6 inhibitors (50) and FGFR1 amplifications have been shown to mediate resistance to CDK4/6 inhibitors (51). Third, while we acknowledge the presence of nongenomic mechanisms (29), genomic alterations are potentially explaining endocrine resistance in a large proportion of patients with ILC. Finally, the genomic landscape of ILC metastases does present some differences with regard to IDC metastases. Some of those, such as ERBB2 and FOXA1 mutations, have been related to endocrine resistance.

C. Marchio reports personal fees from Roche, Bayer, MSD, and Thesaro outside the submitted work. F. Clatot reports personal fees and nonfinancial support from BMS and Merck (outside this work), grants from Astra Zeneca (outside this work), personal fees from Lilly (outside this work), and grants, personal fees, and nonfinancial support from Roche (outside this work) outside the submitted work. A. Vincent-Salomon reports personal fees from Roche, AstraZeneca, and Daichi outside the submitted work. C. Sotiriou reports grants from Fondation Contre le Cancer, Breast Cancer Research Foundation, FNRS, and Les Amis de Bordet during the conduct of the study. No potential conflicts of interest were disclosed by the other authors.

F. Richard: Data curation, formal analysis, investigation, writing-original draft, writing-review and editing. S. Majjaj: Formal analysis, investigation. D. Venet: Formal analysis, methodology, writing-review and editing. F. Rothé: Supervision, investigation. J. Pingitore: Data curation. B. Boeckx: Data curation, formal analysis. C. Marchio: Resources, data curation, writing-review and editing. F. Clatot: Data curation, writing-review and editing. F. Bertucci: Resources, data curation, writing-review and editing. O. Mariani: Resources, data curation. C. Galant: Resources, data curation. G. van den Eynden: Resources, data curation, formal analysis. R. Salgado: Resources, data curation, formal analysis. E. Biganzoli: Supervision, methodology, writing-review and editing. D. Lambrechts: Supervision, methodology, writing-review and editing. A. Vincent-Salomon: Resources, data curation, writing-review and editing. G. Pruneri: Resources, data curation. D. Larsimont: Resources, data curation, writing-review and editing. C. Sotiriou: Conceptualization, resources, data curation, supervision, investigation, methodology, project administration, writing-review and editing. C. Desmedt: Conceptualization, resources, data curation, supervision, funding acquisition, investigation, visualization, methodology, writing-original draft, project administration, writing-review and editing.

This work has been supported by “Les Amis de Bordet,” the “Fondation contre le cancer” (FAF-C/2016/762), “The Breast Cancer Research Foundation,” and the Belgian Fonds National de la Recherche Scientifique (F.R.S-FNRS). F. Richard is supported by a grant from the “Fondation Cancer Luxembourg” (FC/2018/07). C. Marchio was supported in part by a grant from the Mayent-Rothschild Foundation to Institut Curie.

The authors thank the patients and their families, the Biobanks from all participating hospitals, David Brown, Sophia Leduc, Imane Bachir, Yacine Bareche, Marion Maetens, Floriane Dupont, Dominique Roels, and Jeanne Letor for their technical and clinical support.

The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.

1.
McCart Reed
AE
,
Kutasovic
JR
,
Lakhani
SR
,
Simpson
PT
. 
Invasive lobular carcinoma of the breast: morphology, biomarkers and 'omics
.
Breast Cancer Res
2015
;
17
:
12
.
2.
Pestalozzi
BC
,
Zahrieh
D
,
Mallon
E
,
Gusterson
BA
,
Price
KN
,
Gelber
RD
, et al
Distinct clinical and prognostic features of infiltrating lobular carcinoma of the breast: combined results of 15 International Breast Cancer Study Group clinical trials
.
J Clin Oncol
2008
;
26
:
3006
14
.
3.
Mathew
A
,
Rajagopal
PS
,
Villgran
V
,
Sandhu
GS
,
Jankowitz
RC
,
Jacob
M
, et al
Distinct pattern of metastases in patients with invasive lobular carcinoma of the breast
.
Geburtshilfe Frauenheilkd
2017
;
77
:
660
6
.
4.
Arpino
G
,
Bardou
VJ
,
Clark
GM
,
Elledge
RM
. 
Infiltrating lobular carcinoma of the breast: tumor characteristics and clinical outcome
.
Breast Cancer Res
2004
;
6
:
R149
56
.
5.
Desmedt
C
,
Zoppoli
G
,
Gundem
G
,
Pruneri
G
,
Larsimont
D
,
Fornili
M
, et al
Genomic characterization of primary invasive lobular breast cancer
.
J Clin Oncol
2016
;
34
:
1872
81
.
6.
Ciriello
G
,
Gatza
ML
,
Beck
AH
,
Wilkerson
MD
,
Rhie
SK
,
Pastore
A
, et al
Comprehensive molecular portraits of invasive lobular breast cancer
.
Cell
2015
;
163
:
506
19
.
7.
Michaut
M
,
Chin
SF
,
Majewski
I
,
Severson
TM
,
Bismeijer
T
,
de Koning
L
, et al
Integration of genomic, transcriptomic and proteomic data identifies two biologically distinct subtypes of invasive lobular breast cancer
.
Sci Rep
2016
;
6
:
18517
.
8.
Pereira
B
,
Chin
SF
,
Rueda
OM
,
Vollan
HK
,
Provenzano
E
,
Bardwell
HA
, et al
The somatic mutation profiles of 2,433 breast cancers refines their genomic and transcriptomic landscapes
.
Nat Commun
2016
;
7
:
11479
.
9.
Desmedt
C
,
Salgado
R
,
Fornili
M
,
Pruneri
G
,
Van den Eynden
G
,
Zoppoli
G
, et al
Immune infiltration in invasive lobular breast cancer
.
J Natl Cancer Inst
2018
;
110
:
768
76
.
10.
Bajrami
I
,
Marlow
R
,
van de Ven
M
,
Brough
R
,
Pemberton
HN
,
Frankum
J
, et al
E-Cadherin/ROS1 inhibitor synthetic lethality in breast cancer
.
Cancer Discov
2018
;
8
:
498
515
.
11.
Du
T
,
Zhu
L
,
Levine
KM
,
Tasdemir
N
,
Lee
AV
,
Vignali
DAA
, et al
Invasive lobular and ductal breast carcinoma differ in immune response, protein translation efficiency and metabolism
.
Sci Rep
2018
;
8
:
7205
.
12.
Tille
JC
,
Vieira
AF
,
Saint-Martin
C
,
Djerroudi
L
,
Furhmann
L
,
Bidard
FC
, et al
Tumor-infiltrating lymphocytes are associated with poor prognosis in invasive lobular breast carcinoma
.
Mod Pathol
2020
.
DOI: 10.1038/s41379-020-0561-9.
13.
Razavi
P
,
Chang
MT
,
Xu
G
,
Bandlamudi
C
,
Ross
DS
,
Vasan
N
, et al
The genomic landscape of endocrine-resistant advanced breast cancers
.
Cancer Cell
2018
;
34
:
427
38.e6
.
14.
Bertucci
F
,
Ng
CKY
,
Patsouris
A
,
Droin
N
,
Piscuoglio
S
,
Carbuccia
N
, et al
Genomic characterization of metastatic breast cancers
.
Nature
2019
;
569
:
560
4
.
15.
Wagle
N
,
Painter
C
,
Anastasio
E
,
Dunphy
M
,
McGillicuddy
M
,
Kim
D
, et al
The Metastatic Breast Cancer (MBC) project: accelerating translational research through direct patient engagement
.
J Clin Oncol
35
:
15s
, 
2018
(
suppl; abstr 1076
).
16.
Angus
L
,
Smid
M
,
Wilting
SM
,
van Riet
J
,
Van Hoeck
A
,
Nguyen
L
, et al
The genomic landscape of metastatic breast cancer highlights changes in mutation and signature frequencies
.
Nat Genet
2019
;
51
:
1450
8
.
17.
Sokol
ES
,
Feng
YX
,
Jin
DX
,
Basudan
A
,
Lee
AV
,
Atkinson
JM
, et al
Loss of function of NF1 is a mechanism of acquired resistance to endocrine therapy in lobular breast cancer
.
Ann Oncol
2018
;
30
:
115
23
.
18.
Desmedt
C
,
Pingitore
J
,
Rothé
F
,
Marchio
C
,
Clatot
F
,
Rouas
G
, et al
ESR1 mutations in metastatic lobular breast cancer patients
.
NPJ Breast Cancer
2019
;
5
:
9
.
19.
Salgado
R
,
Denkert
C
,
Demaria
S
,
Sirtaine
N
,
Klauschen
F
,
Pruneri
G
, et al
The evaluation of tumor-infiltrating lymphocytes (TILs) in breast cancer: recommendations by an International TILs Working Group 2014
.
Ann Oncol
2014
;
26
:
259
71
.
20.
Li
H
,
Durbin
R
. 
Fast and accurate long-read alignment with Burrows-Wheeler transform
.
Bioinformatics
2010
;
26
:
589
95
.
21.
Chen
X
,
Schulz-Trieglaff
O
,
Shaw
R
,
Barnes
B
,
Schlesinger
F
,
Källberg
M
, et al
Manta: rapid detection of structural variants and indels for germline and cancer sequencing applications
.
Bioinformatics
2016
;
32
:
1220
2
.
22.
Kim
S
,
Scheffler
K
,
Halpern
AL
,
Bekritsky
MA
,
Noh
E
,
Källberg
M
, et al
Strelka2: fast and accurate calling of germline and somatic variants. Nature Methods
.
Nature Publishing Group
2018
;
15
:
591
4
.
23.
Nilsen
G
,
Liestøl
K
,
Van Loo
P
,
Moen Vollan
HK
,
Eide
MB
,
Rueda
OM
, et al
Copynumber: efficient algorithms for single- and multi-track copy number segmentation
.
BMC Genomics
2012
;
13
:
591
.
24.
Carter
SL
,
Cibulskis
K
,
Helman
E
,
McKenna
A
,
Shen
H
,
Zack
T
, et al
Absolute quantification of somatic DNA alterations in human cancer
.
Nat Biotechnol
2012
;
30
:
413
21
.
25.
Nik-Zainal
S
,
Davies
H
,
Staaf
J
,
Ramakrishna
M
,
Glodzik
D
,
Zou
X
, et al
Landscape of somatic mutations in 560 breast cancer whole-genome sequences
.
Nature
2016
;
534
:
47
54
.
26.
Magnani
L
,
Frigè
G
,
Gadaleta
RM
,
Corleone
G
,
Fabris
S
,
Kempe
H
, et al
Acquired CYP19A1 amplification is an early specific mechanism of aromatase inhibitor resistance in ERα metastatic breast cancer
.
Nat Genet
2017
;
49
:
444
50
.
27.
Cerami
E
,
Gao
J
,
Dogrusoz
U
,
Gross
BE
,
Sumer
SO
,
Aksoy
BA
, et al
The cBio Cancer Genomics Portal: an open platform for exploring multidimensional cancer genomics data
.
Cancer Discov
2012
;
2
:
401
4
.
28.
Shiino
S
,
Kinoshita
T
,
Yoshida
M
,
Jimbo
K
,
Asaga
S
,
Takayama
S
, et al
Prognostic impact of discordance in hormone receptor status between primary and recurrent sites in patients with recurrent breast cancer
.
Clin Breast Cancer
2016
;
16
:
e133
40
.
29.
Hanker
AB
,
Sudhan
DR
,
Arteaga
CL
. 
Overcoming endocrine resistance in breast cancer
.
Cancer Cell
2020
;
37
:
496
513
.
30.
Turner
N
,
Kingston
B
,
Kilburn
L
,
Kernaghan
S
,
Wardley
AM
,
Macpherson
I
, et al
Results from the plasmaMATCH trial: a multiple parallel cohort, multi-centre clinical trial of circulating tumour DNA testing to direct targeted therapies in patients with advanced breast cancer (CRUK/15/010) [abstract]
.
In
:
Proceedings of the 2019 San Antonio Breast Cancer Symposium; 2019 Dec 10–14; San Antonio, TX
.
Philadelphia (PA)
:
AACR
;
Cancer Res 2020;80(4 Suppl):Abstract nr GS3-06
.
31.
Xu
G
,
Chhangawala
S
,
Cocco
E
,
Razavi
P
,
Cai
Y
,
Otto
JE
, et al
ARID1A determines luminal identity and therapeutic response in estrogen-receptor-positive breast cancer
.
Nat Genet
2020
;
52
:
198
207
.
32.
Nayar
U
,
Cohen
O
,
Kapstad
C
,
Cuoco
MS
,
Waks
AG
,
Wander
SA
, et al
Acquired HER2 mutations in ER+ metastatic breast cancer confer resistance to estrogen receptor-directed therapies
.
Nat Genet
2019
;
51
:
207
16
.
33.
Jeselsohn
R
,
Yelensky
R
,
Buchwalter
G
,
Frampton
G
,
Meric-Bernstam
F
,
Gonzalez-Angulo
AM
, et al
Emergence of constitutively active estrogen receptor-alpha mutations in pretreated advanced estrogen receptor-positive breast cancer
.
Clin Cancer Res
2014
;
20
:
1757
67
.
34.
Jeselsohn
R
,
De Angelis
C
,
Brown
M
,
Schiff
R
. 
The evolving role of the estrogen receptor mutations in endocrine therapy-resistant breast cancer
.
Curr Oncol Rep
2017
;
19
:
35
.
35.
Fu
X
,
Creighton
CJ
,
Biswal
NC
,
Kumar
V
,
Shea
M
,
Herrera
S
, et al
Overcoming endocrine resistance due to reduced PTEN levels in estrogen receptor-positive breast cancer by co-targeting mammalian target of rapamycin, protein kinase B, or mitogen-activated protein kinase kinase
.
Breast Cancer Res
2014
;
16
:
430
.
36.
Pearson
A
,
Proszek
P
,
Pascual
J
,
Fribbens
C
,
Shamsher
MK
,
Kingston
B
, et al
Inactivating NF1 mutations are enriched in advanced breast cancer and contribute to endocrine therapy resistance
.
Clin Cancer Res
2020
;
26
:
608
22
.
37.
Giltnane
JM
,
Hutchinson
KE
,
Stricker
TP
,
Formisano
L
,
Young
CD
,
Estrada
MV
, et al
Genomic profiling of ER+ breast cancers after short-term estrogen suppression reveals alterations associated with endocrine resistance
.
Sci Transl Med
2017
;
9
:
eaai7993
.
38.
Nagle
AM
,
Levine
KM
,
Tasdemir
N
,
Scott
JA
,
Burlbaugh
K
,
Kehm
J
, et al
Loss of E-cadherin enhances IGF1-IGF1R pathway activation and sensitizes breast cancers to anti-IGF1R/InsR inhibitors
.
Clin Cancer Res
2018
;
24
:
5165
77
.
39.
André
F
,
Ciruelos
E
,
Rubovszky
G
,
Campone
M
,
Loibl
S
,
Rugo
HS
, et al
Alpelisib for PIK3CA-mutated, hormone receptor-positive advanced breast cancer
.
N Engl J Med
2019
;
380
:
1929
40
.
40.
Teo
K
,
Gómez-Cuadrado
L
,
Tenhagen
M
,
Byron
A
,
Rätze
M
,
van Amersfoort
M
, et al
E-cadherin loss induces targetable autocrine activation of growth factor signalling in lobular breast cancer
.
Sci Rep
2018
;
8
:
15454
.
41.
Juric
D
,
Castel
P
,
Griffith
M
,
Griffith
OL
,
Won
HH
,
Ellis
H
, et al
Convergent loss of PTEN leads to clinical resistance to a PI(3)Kalpha inhibitor
.
Nature
2014
;
518
:
240
4
.
42.
Hoadley
KA
,
Siegel
MB
,
Kanchi
KL
,
Miller
CA
,
Ding
L
,
Zhao
W
, et al
Tumor evolution in two patients with basal-like breast cancer: a retrospective genomics study of multiple metastases
.
PLoS Med
2016
;
13
:
e1002174
.
43.
Savas
P
,
Teo
ZL
,
Lefevre
C
,
Flensburg
C
,
Caramia
F
,
Alsop
K
, et al
The subclonal architecture of metastatic breast cancer: results from a prospective community-based rapid autopsy program “CASCADE
PLoS Med
2016
;
13
:
e1002204
.
44.
Brown
D
,
Smeets
D
,
Székely
B
,
Larsimont
D
,
Szász
AM
,
Adnet
P-Y
, et al
Phylogenetic analysis of metastatic progression in breast cancer using somatic mutations and copy number aberrations
.
Nat Commun
2017
;
8
:
14944
.
45.
Avigdor
BE
,
Cimino-Mathews
A
,
DeMarzo
AM
,
Hicks
JL
,
Shin
J
,
Sukumar
S
, et al
Mutational profiles of breast cancer metastases from a rapid autopsy series reveal multiple evolutionary trajectories
.
JCI Insight
2017
;
2
:
e96896
.
46.
De Mattos-Arruda
L
,
Sammut
S-J
,
Ross
EM
,
Markowetz
F
,
Seoane
J
,
Caldas Correspondence
C
. 
The genomic and immune landscapes of lethal metastatic breast cancer
.
Cell Rep
2019
;
27
:
2690
708.e10
.
47.
Siegel
MB
,
He
X
,
Hoadley
KA
,
Hoyle
A
,
Pearce
JB
,
Garrett
AL
, et al
Integrated RNA and DNA sequencing reveals early drivers of metastatic breast cancer
.
J Clin Invest
2018
;
128
:
1371
83
.
48.
Levine
KM
,
Priedigkeit
N
,
Basudan
A
,
Tasdemir
N
,
Sikora
MJ
,
Sokol
ES
, et al
FGFR4 overexpression and hotspot mutations in metastatic er+ breast cancer are enriched in the lobular subtype
.
NPJ Breast Cancer
2019
;
5
:
19
.
49.
Du
T
,
Sikora
MJ
,
Levine
KM
,
Tasdemir
N
,
Riggins
RB
,
Wendell
SG
, et al
Key regulators of lipid metabolism drive endocrine resistance in invasive lobular breast cancer
.
Breast Cancer Res
2018
;
20
:
106
.
50.
Costa
C
,
Wang
Y
,
Ly
A
,
Hosono
Y
,
Ellen
M
,
Walmsley
CS
, et al
PTEN loss mediates clinical cross-resistance to CDK4/6 and PI3Kα inhibitors in breast cancer
.
Cancer Discov
2020
;
10
:
72
85
.
51.
Formisano
L
,
Lu
Y
,
Servetto
A
,
Hanker
AB
,
Jansen
VM
,
Bauer
JA
, et al
Aberrant FGFR signaling mediates resistance to CDK4/6 inhibitors in ER+ breast cancer
.
Nat Commun
2019
;
10
:
1373
.