Article Text

Download PDFPDF

Hereditary lobular breast cancer with an emphasis on E-cadherin genetic defect
  1. Giovanni Corso1,
  2. Joana Figueiredo2,3,
  3. Carlo La Vecchia4,
  4. Paolo Veronesi1,5,
  5. Gabriella Pravettoni5,6,
  6. Debora Macis7,
  7. Rachid Karam8,
  8. Roberto Lo Gullo9,
  9. Elena Provenzano10,11,12,
  10. Antonio Toesca1,
  11. Ketti Mazzocco5,6,
  12. Fátima Carneiro3,13,
  13. Raquel Seruca2,3,14,
  14. Soraia Melo2,3,14,
  15. Fernando Schmitt2,3,14,
  16. Franco Roviello15,
  17. Alessandra Margherita De Scalzi1,
  18. Mattia Intra1,
  19. Irene Feroce7,
  20. Elisa De Camilli16,
  21. Maria Grazia Villardita1,
  22. Chiara Trentin9,
  23. Francesca De Lorenzi17,
  24. Bernardo Bonanni7,
  25. Viviana Galimberti1
  1. 1 Division of Breast Surgery, European Institute of Oncology, Milano, Italy
  2. 2 EPIC Lab, Instituto de Investigação e Inovação em Saúde (i3S), University of Porto, Porto, Portugal
  3. 3 Institute of Molecular Pathology and Immunology, University of Porto (IPATIMUP), Porto, Portugal
  4. 4 Department of Clinical Sciences and Community Health, University of Milan, Milan, Italy
  5. 5 Oncology and Hematology, University of Milan, Milan, Italy
  6. 6 Applied Research Division for Cognitive and Psychological Science, European Institute of Oncology, Milan, Italy
  7. 7 Division of Cancer Prevention and Genetics, European Institute of Oncology, Milan, Italy
  8. 8 Ambry Genetics, Aliso Viejo, California, USA
  9. 9 Division of Breast Imaging, European Institute of Oncology, Milan, Italy
  10. 10 NIHR Cambridge Biomedical Research Centre, Addenbrooke’s Hospital, Cambridge, UK
  11. 11 Cambridge Breast Cancer Research Unit, Cambridge University Hospitals NHS Foundation Trust, Cambridge, UK
  12. 12 Department of Histopathology, Addenbrooke’s Hospital, Cambridge University Hospitals NHS Foundation Trust, Cambridge, UK
  13. 13 Division of Pathology, Hospital São Joao, Porto, Portugal
  14. 14 Medical Faculty of the University of Porto, Porto, Portugal
  15. 15 Departments of Surgery and Pathology, Le Scotte Hospital, University of Siena, Siena, Italy
  16. 16 Division of Pathology, European Institute of Oncology, Milan, Italy
  17. 17 Division of Plastic Surgery, European Institute of Oncology, Milan, Italy
  1. Correspondence to Dr Giovanni Corso, Division of Breast Surgery, European Institute of Oncology, Milano 20141, Italy; giovanni.corso{at}


Recent studies have reported germline CDH1 mutations in cases of lobular breast cancer (LBC) not associated with the classical hereditary diffuse gastric cancer syndrome. A multidisciplinary workgroup discussed genetic susceptibility, pathophysiology and clinical management of hereditary LBC (HLBC). The team has established the clinical criteria for CDH1 screening and results’ interpretation, and created consensus guidelines regarding genetic counselling, breast surveillance and imaging techniques, clinicopathological findings, psychological and decisional support, as well as prophylactic surgery and plastic reconstruction. Based on a review of current evidence for the identification of HLBC cases/families, CDH1 genetic testing is recommended in patients fulfilling the following criteria: (A) bilateral LBC with or without family history of LBC, with age at onset <50 years, and (B) unilateral LBC with family history of LBC, with age at onset <45 years. In CDH1 asymptomatic mutant carriers, breast surveillance with clinical examination, yearly mammography, contrast-enhanced breast MRI and breast ultrasonography (US) with 6-month interval between the US and the MRI should be implemented as a first approach. In selected cases with personal history, family history of LBC and CDH1 mutations, prophylactic mastectomy could be discussed with an integrative group of clinical experts. Psychodecisional support also plays a pivotal role in the management of individuals with or without CDH1 germline alterations. Ultimately, the definition of a specific protocol for CDH1 genetic screening and ongoing coordinated management of patients with HLBC is crucial for the effective surveillance and early detection of LBC.

  • lobular breast cancer
  • E-cadherin
  • hereditary cancer

Statistics from


Hereditary lobular breast cancer (HLBC) is a rare inherited predisposition associated with CDH1 germline mutation, without apparent correlation with hereditary diffuse gastric cancer (HDGC).1

CDH1 germline mutations are responsible for HDGC syndrome development.2 Over 15 years ago, the first germline mutations were identified in a Maori population from New Zealand with a strong cluster for gastric tumours.3 Initially, this ethnic group also showed a cluster with women affected by primary breast tumour, therefore the International Gastric Cancer Linkage Consortium (IGCLC) defined LBC cases carrying any pathogenic CDH1 germline mutation as an additional clinical outcome of the HDGC syndrome.4 In particular, the IGCLC recognised that heterozygous germline CDH1 mutations increase lifetime risk of developing gastric tumours and breast tumours, based on the fact that LBCs were always identified in families with aggregation of DGCs.

However, recent research studies identified novel CDH1 germline variants with a potential pathogenic role in women with diagnosed LBC (in invasive and/or in situ patterns) and without family history of gastric carcinoma.5 This was the first evidence of a novel independent inherited cancer predisposition.

Given the lack of knowledge and the need for specific guidelines for HLBC, an international consensus workgroup of surgeons, pathologists, clinical geneticists, molecular biologists, breast radiologists and epidemiologists with high experience in this field was assembled in order to define management of patients affected by this cancer syndrome. This report is the result of the first HLBC consortium. Herein, we describe the current evidence and HLBC-oriented recommendations considering specific sections: (A) epidemiology, (B) CDH1 germline variants, (C) pathology, (D) genetic counselling, (E) breast imaging and surveillance, (F) psycho-oncology and prophylactic surgery with breast reconstruction. Additionally, the Clinical Genome Resource (ClinGen) Network convened the ClinGen Expert Panel to future clarify the role of CDH1 germline variants in disease context.


In the European Union (EU), BC mortality (world standard) declined from 17.9/100 000 in 2002 to 15.2 in 2012. The predicted 2020 rate is 13.4/100 000. The decline was largest in young women. Over 30 000 BC deaths will be avoided in 2020 in the EU as compared with the peak rate in 1989, with a total of 450 000 BC deaths avoided over the period 1990–2020.6 7

BC mortality trends were favourable in North America and Oceania over the last three decades, and a further 10% reduction in the overall rates in these regions is predicted for 2020, to reach values of 11–12/100 000 women, that is, about half the top rates in the late 1980s.8

About 10% of all BCs are lobular carcinomas. BRCA2 mutations are associated with both ductal and LBC, while CDH1, the gene coding for the E-cadherin adhesion protein, is specifically related with invasive lobular cancer (ILC).9

The lifetime probability of developing BC rises from about 10%–12% in the general population, to over 30% for a woman with early-onset family history and high lifestyle profile and over 60% for BRCA1 and BRCA2 mutant carriers.7 10 11 In particular, women with LBC and family history of cancer more commonly have a father than a mother diagnosed with cancer. In a clinical series of 1676 breast tumours from Sweden, including 141 lobular breast tumours, the relative risk (RR) was about twofold elevated in patients with a family history of cancer, and remained similar after excluding prostate cancer.12

Strong familial and genetic factors are however restricted to a minority of women, including in those with LBC.

There is currently inadequate information to distinguish specific risk factors between ductal and LBCs, thus in the following section, we will review a number of key risk factors related to the epidemiology of BC overall.13

Reproductive, menstrual and hormonal factors

BC incidence and risk is associated with menstrual and reproductive factors related to female hormones.13

BC risk is inversely related to parity, age at menarche and directly to age at first birth and at menopause. The excess breast tumour risk for each year of delay of menopause (about 2.5%) is similar to that for 1 year use of menopause hormone replacement therapy (HRT).

BC risk is also directly related to being overweight postmenopause, and this is due to increased oestrogen levels and availability in postmenopausal women. Avoidance of long-term HRT use, and weight loss and reduction of obesity could thus avoid 15%–20% of postmenopausal BCs.

Timing of endogenous hormone exposure is difficult to define, hence the time-risk relationship is poorly described—although long-term excess weight and obesity are associated to greater BC risk, while the RR levels off in women who lose weight.

Current and recent use of oral contraceptives (OC) and HRT (combination oestrogen–progestin HRT) is also related to increased BC risk. However, the RR declines in the short term after stopping OC and HRT use, reducing the public health implications of this issue.

The RR for alcohol is increased by about 5% per drink per day. Thus, reduction of drinking could avoid about 5% of BCs in Europe. A favourable influence of physical activity and selected aspects of diet on BC risk is also possible, though significance of these factors remains uncertain. Still, up to 10% of BCs could be avoided by an adequate level of physical activity. A similar proportion could be avoided by a diet rich in vegetables and poor in animal fats, assuming an RR of 0.8 for a favourable diet.14 In a case–control study conducted in Italy and Switzerland including 3034 BC cases and 3392 controls, adherence to the Mediterranean diet was measured through a Mediterranean Diet Score (MDS) summarising the major characteristics of the dietary pattern and ranging from 0 (lowest adherence) to 9 (highest adherence). Compared with MDS of 0–3, the ORs of breast tumour were 0.86 (95% CI 0.76 to 0.98) for MDS of 4–5 and 0.82 for MDS of 6–9.15 16

Lactation has a protective effect on BC, but given the limited number of births per woman in Europe, lactation has a modest impact on lifetime BC risk. Likewise, BC risk could be reduced by earlier first birth and increased number of births, but these factors imply complex societal changes.

Thus, modification of selected lifestyle factors would have an appreciable impact in reducing the incidence of BC in Europe. At an individual level, the risk reduction would be larger in women with a family history or hereditary breast cancer, including those with HLBC.

E-cadherin germline variants

Classically, patients with LBC carrying CDH1 germline mutations are associated with the HDGC syndrome.17 Indeed, early-onset LBC might be the first presentation of HDGC. Benusiglio et al identified CDH1 germline deleterious mutations in three bilateral LBC cases (age at onset <50 years) not fulfilling the IGCLC criteria. These families were negative at the beginning for HDGC in first and second-degree relatives and without BRCA1 and BRCA2 alterations.5 CDH1 mutations have been identified in four bilateral early-onset LBCs (age at onset <50 years) with no family history of HDGC.18 Furthermore, Silva et al reported two female patients with LBC (ages 60 and 51), both carriers of germline CDH1 mutations, who underwent prophylactic gastrectomy that revealed foci of intramucosal DGC in both cases.19

Importantly, very recently two potentially pathogenic CDH1 alterations were reported in patients with LBC, and without family history of gastric tumour.20 Considering only individuals with LBC and excluding pedigree with gastric tumour manifestation, a total of 495 female LBCs from six independent original studies were screened for CDH1 genetic alterations. Sixteen (3.2%) novel CDH1 germline variants have been identified (table 1). Bilateral LBC was diagnosed in 14.3% of cases (6/63), positive family history of BC was reported in 40.3% (163/404) and the mean age at onset was about 45 years old.

Table 1

Identified CDH1 germline alterations in families with lobular breast carcinoma and without aggregation for diffuse gastric cancers

The type and frequency of identified CDH1 variants occurred as follows: six missense (37.5%), three splice site (18.7%), three deletions (18.7%), two insertions (13%), one non-sense (6.2%) and one stop codon (6.2%) (table 1). These variants affect different CDH1 gene domains, spanning almost all 16 exons and 1-7-13 intron boundaries (figure 1).

Figure 1

Distribution of CDH1 germline variants identified in the hereditary lobular breast cancer (HLBC) context. The location of all CDH1 germline variants found in patients with HLBC is represented in an illustration of E-cadherin. The protein structure comprises the signal peptide, precursor, extracellular domain, transmembrane domain (TM) and the cytoplasmic domain.

In accordance with the American College of Medical Genetics and Genomics (ACMG) guidelines, germline alterations should be classified as pathogenic and benign variants. Variant of uncertain significance (VUS) is a nomenclature to classify alteration with an as yet undefined pathogenic role.

The pathogenic role of these CDH1 variants was discussed with the ClinGen. We verified that P30T, A592T and V832M alterations are classified as ‘benign’, since they are found respectively in 0.25%, 0.48% and 0.25% in the gnomAD European and East Asian cohorts. Conversely, alterations A408T, D443N and S838G are classified as VUS (table 1).

The unsolved enigma of missense mutations

In the LBC setting, missense mutations constitute the most frequent CDH1 genetic alterations. This mutation type preserves full-length protein but induces an amino acid substitution, which can affect E-cadherin structure or docking sites for a number of direct interactors, compromising E-cadherin processing, localisation and function. Mutations with such deleterious impact are classified as pathogenic, while those not altering E-cadherin properties are considered polymorphisms.21 Given the possibility of a functional or a dysfunctional result/protein, the classification of CDH1 missense variants constitutes a serious burden for counselling and surveillance of mutant carriers.22 23

To assess the functional relevance of CDH1 missense variants in families with HLBC, a similar protocol to that established for HDGC cases should be employed. An integrative workflow encompassing the analysis of familial and population data, in silico analysis and functional in vitro assays has been recommended.4 21 22

The cosegregation of the mutation with the disease should first be analysed for the identification of inheritance patterns in the family.1 21 23 For that purpose, several family members need to be screened for the presence of CDH1 alterations.4 21 22 However, geneticists are frequently confronted with families of small size and with a low number of affected members, which prevents segregation analysis.22 Mutation recurrence in unrelated CDH1 families and variant frequency in a control or a general population should also be evaluated.4 21 22

Publicly available population databases such as 1000 Genomes Project (, Exome Variant Server ( or dbSNP Database ( can be instrumental for this analysis.21

In silico tools have been applied to evaluate putative impact of variants on the primary and alternative gene transcripts, and also on the protein structure and function. Predictions are based on the conservation of a particular nucleotide among species, the location and function of the affected region, the biochemical properties of the amino acid substitution, the protein native-state stability and the possible effect in splice sites.22 24 Thus, the theoretical tools most frequently used to determine the value of CDH1 missense mutations are SIFT—Sorting Intolerant from Tolerant (, PolyPhen-2—Polymorphism Phenotyping v2 (, FoldX ( and Netgene2 algorithm ( 24 25 Despite the useful information gathered with bioinformatics tools, clinicians and researchers should be aware of their limitations concerning the lack of possible compensatory effects of neighbouring positions and the scarce transcriptional and structural data available.24 26

In vitro assays were implemented to test the effects of variants in protein expression levels, protein localisation, cell-cell adhesion and invasive patterns.21 27 For the experiments, cell lines are transfected with vectors encoding the variants and the wild-type protein, as a control.24 27 28 Chinese hamster ovary cells are the conventionally accepted model to perform all the tests as this cell line is completely negative for cadherin expression and, on transfection with the wild-type E-cadherin, the cells gain the capacity to form cellular aggregates and to suppress invasion through artificial extracellular matrices.27 28 Total E-cadherin expression levels must be assessed to look for structural destabilisation, premature degradation of the protein by the mechanisms of protein quality control, or abnormal glycosylation patterns.24 29 30 The study of E-cadherin localisation by immunostaining is also crucial to identify deleterious variants. In particular, protein diffusely distributed throughout the cell or abnormally accumulated in cytoplasmic organelles is a strong indication of E-cadherin loss of function.28 29 31 Nevertheless, some pathogenic variants are able to reach the membrane but not to mediate a stable cell-cell contact.28 31

The ability of a mutant protein to establish homophilic binding with E-cadherin molecules on neighbouring cells is evaluated through slow aggregation assays.27 28 32 Specifically, cells expressing a functional variant assemble a competent adhesion complex and spontaneously aggregate on a semisolid agar substrate.33 In contrast, cells presenting dysfunctional E-cadherin forms could generate small cellular aggregates or present an isolated phenotype.27 28 32–34

Cellular distribution patterns elicited by E-cadherin variants were recently proposed as a complementary technique to investigate cell–cell interactions. The approach involves the generation and quantitative analysis of networks that precisely represent cell organisation. It was verified that the connection of dysfunctional E-cadherin cells yields loose networks with cells more distant from each other, corroborating a model of poorly cohesive epithelial structures.35

Invasive properties of cells expressing novel variants are examined by using Matrigel invasion chambers, which epitomise the heterogeneous composition of the basement membrane in vitro.36 37 Accordingly, neutral variants maintain the ability to suppress invasion, while the pathogenic forms lose this function.27–29

To date, a large number of CDH1 germline mutations were found in different clinical settings and reported for functional evaluation at the Seruca’s lab (IPATIMUP/i3S, Porto).24 34 The vast majority of them (60.7%) demonstrated to impair E-cadherin function in vitro, 24.6% showed a similar performance to that of the wild-type form and 14.8% of variants remain inconclusive. For the latter, complementary experiments such as the analysis of cell migration, the interplay between E-cadherin and its direct binding partners, as well as the study of downstream targets could be applied.27 28 38 39

The ClinGen Network

In recent years, sequencing technology has evolved rapidly with the advent of high-throughput next-generation sequencing (NGS). The implementation of NGS has resulted in a dramatic increase in the identification of germline variants by clinical diagnostic laboratories.40 The increased availability of genetic testing has also been accompanied by new challenges in sequence interpretation.41

In this context, the ACMG developed guidelines for the interpretation of germline sequence variants.42 These recommendations describe the process for classifying variants into five categories—‘pathogenic’, ‘likely pathogenic’, ‘uncertain significance’, ‘likely benign’ and ‘benign’—based on criteria using typical types of variant evidence (eg, population data, computational data, functional data, segregation data). However, the ACMG guidelines often do not take into consideration gene-specific nuances, such as disease-specific incidence and prevalence rates.

With this in mind, the National Institutes of Health funded the ClinGen Network, an expert panel dedicated to building an authoritative central resource that defines the clinical relevance of genes and variants for use in precision medicine and research.43 Several key goals support ClinGen’s overall mission, including optimisation of gene-specific clinical annotation and interpretation of genomic variants, and sharing of genomic data through a centralised database for clinical and research use ( In the case of CDH1, the ClinGen Expert Panel was assembled with the primary goal of optimising the 2015 ACMG/AMP Variant Interpretation Guidelines, and to make specific recommendations for the assessment of CDH1 germline variants.

Pathology of LBC

ILC was first described by Foote and Stewart in 1941 and,44 subsequently, in the 1970s and 1980s several distinct variants of ILC were reported, including the alveolar, solid, pleomorphic, signet ring, histiocytoid and apocrine subtypes. ILC is the second most common histotype after (infiltrating) ductal carcinoma.

Overall, ILC and its variants represent 5%–15% of all invasive BCs and demonstrate peculiar clinicopathological features.

Typically, ILC forms a firm-to-hard tumour with irregular borders. The edges of the lesion may be more easily appreciated by palpation than by inspection, because the margin may blend imperceptibly with the surrounding parenchyma. The classic type constituted small to medium-sized, uniform, discohesive tumour cells with mild nuclear atypia and slight hyperchromasia, arranged in linear strands. Occasionally, tumour cells grew around ducts and lobules in a concentric (targetoid) pattern. In the alveolar subtype, tumour cells with the same uniform appearance as in the classic type were clustered in aggregates of >20 cells, separated by thin bands of fibrous stroma. The tumour cells of the solid type cells diffusely infiltrated the surrounding tissues in large solid sheets, with scarce intervening stroma. The pleomorphic type demonstrated architectural features similar to the classic type but the neoplastic cells demonstrated marked cellular pleomorphism and nuclear atypia. Apocrine differentiation and histiocytoid-like appearance were detectable in ILC otherwise recapitulating the cytoarchitectural features of the pleomorphic type, but could represent the predominant cell populations of distinct tumours. Signet ring cells (ie, tumour cells demonstrating intracytoplasmic accumulation of mucins with peripheral dislocation of the nucleus) could represent the exclusive cell component of ILC (signet ring cell lobular carcinoma), but could be found in different subtypes as well. The tubular lobular variant of ILC was characterised by an infiltrative pattern similar to that of classic ILC but with some tumour cells organised in tubular structures of a smaller size than those found in pure tubular carcinoma. At the cytological level, the tumour cells were described as ‘small or medium-sized’, ‘rather uniform in their staining properties’ and as exhibiting relatively little ‘irregularity’. Usually, the tumour cells are oestrogen receptor (ER) and progesterone receptor positive, without overexpression or amplification of the HER-2/neu gene (figure 2).

Figure 2

Invasive lobular carcinoma. Circumferential infiltration around an atrophic lobule is shown. H&E staining (20×) (A). Note membrane reactivity with immunohistochemistry  (IHC) E-cadherin (10×) in the centrally entrapped benign  duct (B).

Lobular breast tumorigenesis with CDH1 defect

Lobular in situ neoplasia (LN), comprising lobular cancer in situ (LCIS) and atypical lobular hyperplasia (ALH), is characterised histologically by a proliferation of small non-cohesive cells within the terminal duct lobular unit. In ALH, the distension involves less than 50% of the acinus, progressing to LCIS where more than 50% of acini are expanded with loss of the central lumina.45 The characteristic appearance is due to loss of expression of E-cadherin, a cell-cell adhesion molecule that forms part of the cadherin-catenin complex that plays a key role in maintenance of lobular architecture.46 47

Detection of LCIS is increasing, and it is identified in 0.5%–3.8% of breast biopsies.48 When present, LCIS is commonly multicentric (>50% of cases) and bilateral (30%).48 Risk of subsequent invasive carcinoma is increased four to five times with ALH and 8 to 10 times with LCIS49; the subsequent invasive cancers may be ductal or lobular in nature and can arise in either breast, so for a long time LN was regarded as an indicator of risk. More recent series show ipsilateral cancers are three times more likely with an over-representation of LBC so LCIS is now understood to be a true non-obligate precursor lesion although with a low risk of progression to invasive disease.50 This is supported by molecular studies that show shared genetic changes in LCIS and invasive lobular breast cancer (ILBC) on analysis with loss of heterozigosity (LOH), comparative genomic hybridization (CGH) and whole genome sequencing.48 51 52

LCIS forms part of the low-grade ER-positive neoplasia pathway associated with chromosomal losses on 16q and 17p and gains on 1q.53 A stepwise model for progression from normal lobules to ALH, LCIS and ILBC is provided in figure 3. Both ALH and LCIS show similar genetic changes with loss of 16q21-23.1, the locus of the CDH1 gene, on array CGH.54 Mutations in CDH1 have been identified in up to 56% of cases of LCIS, with matching CDH1 mutations in adjacent ILBC.52 55 Frequent mutations are also found in PIK3CA and CBFB in LCIS with shared mutations in 71% of coexistent ILBC indicating that the lesions are clonally related.52 Another study by the same group identified 169 genes that were differentially expressed in the progression pathway from normal epithelium to LCIS to ILBC; interestingly there was enrichment for genes located on 16q and 1q, sites of common copy number changes in lobular carcinoma.56 Other factors required for progression from LCIS to invasion include alterations in myoepithelial cells and in the surrounding microenvironment.57

Figure 3

Representative model of lobular breast cancer (LBC) progression in CDH1 mutation carries. In a CDH1 wild-type situation (left panel), normal lobules are well-organised structures characterised by the strong cell-cell adhesion mediated through the homophilic binding of E-cadherin molecules on neighbouring cells. However, the occurrence of a CDH1 mutation can affect E-cadherin function, resulting in decreased cell-cell adhesion and increased cell proliferation, so-called lobular hyperplasia. Subsequently, the occurrence of a second-hit CDH1 inactivation induces the loss of E-cadherin expression and, consequently, alters organisation of the lobule. During this process, abnormal cells emerge and accumulate in the lobules giving rise to lobular intraepithelial neoplasia. Ultimately, cancer cells disrupt the basement membrane and invade surrounding breast tissues—a stage that is classified as invasive lobular carcinoma.

There is very little in the literature on the frequency and role of ALH and LCIS in HLBC secondary to CDH1 germline mutations. In cases of bilateral LCIS/ILBC within the GLACIER study, four of 50 were found to harbour germline CDH1 mutations, with three truncating mutations and one splice site mutation.18 If women under 45 were considered, 28% of patients with bilateral LCIS had underlying germline mutations. A literature review of 408 cases of LCIS/ILBC with no history of HDGC revealed only three germline CDH1 mutations.18 This included one series of 65 cases of LCIS, 17 of which were bilateral, in which no germline CDH1 mutations were identified.58

Genetic counselling

The first step in the genetic counselling process is the collection of personal and family history, defined as the description of the genetic relationships and medical history of a family.59 A family medical history questionnaire is the most common tool used for taking the necessary information. However, several studies have demonstrated that self-reported family history is often inaccurate and review of medical reports by physicians is advisable.60

Accurate family history data collection is a crucial step for the identification of the most appropriate candidate for genetic testing, which in the majority of cases is the member of the family with the earliest age of onset of cancer.

About 10% of invasive BCs are LBC and it is more strongly associated with exposure to female hormones, therefore its incidence is more subject to variation.9

Indeed, in families with just LBC and/or signet ring GC, HLBC, with an emphasis on E-cadherin genetic defect, can be suspected. CDH1 is the related gene associated with ILC, but never with ductal carcinoma. This gene was initially known as the main susceptibility gene for GC of the diffuse type or signet ring, but the excess of BCs of the lobular type in CDH1 families led researchers to identify it also as a susceptibility gene for ILC.61

CDH1 genetic testing is proposed when there is a personal or family history of DGC and LBC (with one diagnosed <50 years of age), early-onset bilateral lobular carcinoma, or family history of multiple lobular carcinomas at a young age (<50 years) in the absence of gastric tumour in the family.4

If hereditary factors are identified or gene testing reveals variations in the CDH1 gene, all relevant family members should be referred for genetic testing to confirm their carrier status. If a mutation is identified, it is mandatory to build a personalised cancer prevention, screening and management programme with a clinically experienced team.

The latest IGCLC consensus established as mandatory clinical criteria for CDH1 genetic screening personal or family history of HDGC and LBC, one diagnosed <50 years. Testing could also be considered in families presenting with bilateral LBC or family history of two or more cases of LBC <50 years.4

However, clinical criteria for E-cadherin genetic test eligibility are currently under revision, since novel CDH1 germline mutations are identified in individuals not fulfilling these classic criteria.

As documented in the above-mentioned sections, recent CDH1 genetic screening results demonstrated that the latest IGCLC clinical criteria are insufficient and should be revised in accordance with the most recent data.

The current consensus approved that clinical criteria for E-cadherin genetic screening associated with LBC can be reconsidered in two different cancer inherited predispositions: (A) LBC in the setting of the HDGC syndrome, and (B) HLBC without aggregation for gastric tumours.

Considering the ‘so-called’ HLBC, without family history of gastric tumour, E-cadherin genetic test should be proposed in the following cases: (A) bilateral LBC with or without family history of LBC, with age at onset <50 years, and (B) unilateral LBC with family history of LBC, with age at onset <45 years. Whenever possible, BRCA1/2 germline mutations should be excluded in both groups, since they are mutually exclusive with CDH1 germline mutations.

Imaging of LBC and breast surveillance in a high-risk population

Tumour density is often less than or equal to that of normal fibroglandular breast tissue.62 63 These histopathological features limit conventional imaging, particularly mammography, in detecting and staging ILC.

Loss of E-cadherin results in fewer, sparser, cancer cells, failure to elicit a desmoplastic response and more low-density fat per unit volume of tumour compared with ductal carcinomas.64 65

Different somatic mechanisms are responsible for this loss of expression identified in breast tumour: (A) somatic CDH1 mutations, (B) loss of heterozygosity, (C) CDH1 promoter hypermethylation, and (D) allelic imbalance.66

LBC is more often multicentric and bilateral (10%–15%), therefore imaging evaluation of the contralateral breast is crucial.

Mammography, ultrasonography (US) and MRI all play important roles, with each modality having its own advantages and limitations.

The sensitivity of mammography for the detection of ILC varies according to breast density and ranges from 64% to 92%.67 68

Mammographic findings include spiculated mass lesions (42%–53%), architectural distortions (14%–16%) and discrete masses/opacities (about 7%) or asymmetrical densities (about 4%), which are sometimes difficult to detect in the absence of previous mammograms.64 69

Microcalcifications are not a feature of pure ILC (present in only 3%–13% of cases) but may be present in mixed tumours, with an associated component of ductal carcinoma in situ, in pleomorphic LCIS or from juxtaposed areas of sclerosing adenosis and apocrine metaplasia.70 71

Because of the low rate or the absence of suspicious calcification and its tendency to be of low opacity, the mammography may be normal in about 30% of cases with a false negative rate ranging from 19% to 66%.72 73

Digital breast tomosynthesis (DBT) is a recent three-dimensional technique based on digital mammography (DM). This technique reduces summation of overlapping breast tissue and improves detection of ILC presenting as an architectural distortion.74 75

Compared with DM, DBT increases conspicuity of ILC, improves assessment of disease extent and the definition of multifocal multicentric or bilateral ILC. Literature shows that increase of detection rate by addition of DBT to DM was higher for ILC (+107%) than for invasive ductal carcinoma (+33%).76

Though DBT has shown to increase lesion conspicuity, it failed to accurately assess tumour size and therefore it should be regarded as an adjunct to DM but not as an alternative to MRI.77

The use of US and MRI as adjuncts to mammography increases sensitivity in the detection of ILC and provides useful information for further management and presurgical planning.78

Ultrasound imaging is more sensitive than mammography in the visualisation of ILC; with this tool, the sensitivity increases up to 98% (range 86%–98%).68

The most common sonographic appearance is that of a heterogeneous hypoechoic mass with posterior acoustic shadowing (68%–84%) or without shadowing (about 20%), shadowing only (13%–17.5%), a well-circumscribed mass (about 4%), infiltrative pattern (about 4%), a focal hyperechoic lesion (1%–5%) or as no abnormality (2%–4%).64 79

Due to its propensity for multicentricity, breast MRI is usually recommended in many countries when histology of a lesion reveals ILC.80

It is more sensitive in detecting ILC and determining the size extent and presence of multiple or contralateral foci.67 81

A recent meta-analysis conducted on 21 studies shows sensitivity ranging from 83% to 100% (mean 93.3%), equal to the overall sensitivity of MRI for malignant lesion of the breast. At MRI evaluation, ILC appears as a mass-like lesion (with irregular or speculated margins) more frequently than as a non-mass lesion (with abnormal enhancement such as regional, ductal, segmental or diffuse enhancement). There are data to support how MRI changes surgical management in 28.3% of cases.67 A recent study (radiology) demonstrated that preoperative breast MRI in patients with ILC is associated with a reduction in the rate of repeat surgery without an increase in the initial or final mastectomy rates.

Although numerous studies have shown that MRI is useful for breast screening in high-risk women with BRCA1, BRCA2 and TP53 gene mutations, to this day, there are no guidelines regarding screening with MRI in women with CDH1 mutations.

Given the high prevalence of ILC in CDH1 germline mutation carriers, and the histopathological and imaging features of these tumours, breast screening in CDH1-mutated patients should be performed annually with DM (possibly with DBT), ultrasound and contrast-enhanced MRI. A 6-month interval between the US and the MRI is preferable but not mandatory.

Psychological management in asymptomatic mutant carriers

Women with an asymptomatic mutation have to face a critical dilemma: what to do with the information obtained by the genetic testing. While genetic testing reduces uncertainty by providing information on biological and molecular characteristics, it also introduces uncertainty on what to do subsequently. This uncertainty has an important impact on how a person builds a mental representation of the problem, and consequently on how she reacts and what choices she makes.

Given the relevance of uncertainty in patient’s empowerment and clinical outcomes and considering the crucial role of patients’ decision-making in becoming empowered the psychological interventions in these patients should be twofold: help them manage the psychological implications of a genetic mutation and the resultant medical interventions,82 improving coping skills, quality of life and decreasing distress by means of psychological support or psychotherapy, and by supporting them in the decision-making process.83 Both interventions are strictly interrelated, since a good management of the patient’s emotional and psychological state has a strong impact on the decision-making process.

Patients’ decisions are often based on heuristics and,84 most of the time, guided by emotions rather than by objective risks and benefits.85 This is exactly the case in patients with a mutation deciding for a risk-reducing mastectomy (RRM), where the emotional distress derived by the risk of having a breast tumour goes beyond evidence and impairs quality of life.86 87

Thus, besides the fact that RRM with or without breast reconstruction reduces BC risk, it is important to consider that they are invasive procedures that can have iatrogenic consequences,88 and in turn may affect the patient’s body image, identity and relational habits.

For this reason, and especially when a proposed option is a surgical intervention on a still healthy breast, psychological management requires investigation of the information possessed by the patient about the proposed options, the surgical procedures and the clinical, psychological and relational implications of potential RRM. The aim is helping people be aware of their own decision-making process and discuss needs, values, preferences and expectation associated with RRM, in order to increase the congruency between psychological and clinical expectations and outcomes.

Prophylactic surgery

Due to the low number of identified CDH1 germline mutations and scarce data regarding the risk of BC development in CDH1 mutant carriers, there is no conventionally accepted protocol for prophylactic surgery. However, in selected cases (figure 4), if the surgical strategy is discussed and is in accordance with patient’s wishes (also considering the uncontrolled fear, which is frequently observed in these situations), a prophylactic bilateral nipple-sparing mastectomy (PNSM) could be performed as defined for BRCA mutation carriers.89

Figure 4

Flow chart for management of patients with hereditary lobular breast cancer (HLBC). Clinical criteria for genetic counselling and CDH1 testing are presented. Overview of recommended procedures for the surveillance of CDH1-positive and negative patients is also represented. IGCLC, International Gastric Cancer Linkage Consortium; LBC, lobular breast cancer; US, ultrasonography; VUS, variant of uncertain significance.

A recent review by Muller et al highlights the extremely low risk of cancer development following breast risk-reducing surgery. Specifically, from the 3716 cases of PNSM, only nine cases (0.2%) of BC exterior to the nipple areola complex (NAC) and one case (0.004%) within the NAC were reported.90

The complication rate associated with PNSM is comparable to that of skin sparing and total mastectomy, although there is an additional risk of postoperative necrosis of the nipple. A recent meta-analysis of nipple-sparing mastectomy (NSM) reported an overall complication rate of 22% and a nipple necrosis rate of 5.9%.91 On the other hand, the aesthetic outcome can be significantly improved by conserving the NAC.92 93

In the mastectomy process, the breast tissue anatomically localised by splitting of the superficial fascia delimited by the large infraclavicular muscle bundle, the midline of the sternum, the front edge of latissimus dorsi and the lower edge of major pectoralis on the sixth rib should be removed en bloc.

The patient must be informed of residual cancer risk from remaining glandular cells even in the most accurate mastectomies. Respecting the limits of the dissection and saving the inframammary folds allow a better breast reconstruction.

Several skin incisions have been proposed in the literature in order to remove en bloc the glandular breast tissue on one hand, and maintaining the entire breast envelope on the other. The upper outer linear incision is most commonly used because it safeguards vascularisation of the NAC and easily removes all the glandular tissue. However, it is not the right cosmetic approach for every patient, and consideration must be given that during healing, the scar could retract the NAC towards the axillary tail. Moreover, there are also non-conventional surgical approaches tailored to the individual patient94 when the classic incision is not possible.95

Technical innovations have already made feasible robotic NSM, and this has been reported as well tolerated, safe and associated with good cosmetic outcome. Endoscopic NSM has been abandoned due to technical limits circumvented by ‘wristed’ instruments of the robot.96 97 The robotic technique reproduces the open approach, and assisted by the accuracy, flexibility and control given by the robot allows the surgeon to perform complete mastectomy through a tiny hidden extramammary incision.98 A randomised clinical trial comparing open NSM and robotic NSM in prophylactic surgery is ongoing ( number NCT03440398).

Breast reconstruction is an important consideration and potential influence on choice of procedure. For this reason, consultation with a plastic surgeon is mandatory before any risk-reducing surgery, providing information about the type of reconstruction and expected results; patient expectations are very high and women are seeking outstanding cosmetic results.

Although the majority of women elect for breast reconstruction after prophylactic mastectomy, the rates of reconstruction vary widely across different countries and single institutions.99 100

In case of risk-reducing surgery on healthy breasts, a more conservative mastectomy is possible and more ‘aesthetic’ surgical incisions may be performed, since the surgical approach is not restricted by oncological considerations. The reconstruction is usually performed at the time of mastectomy and called ‘immediate’.

Options for reconstructions include both prostheses and autologous tissues.101 Many patients prefer autologous tissue reconstruction, where possible, because of the similarity to natural breast, stability of the long-term result and anxiety about use of foreign material/implants. Autologous reconstruction encompasses a broad range of procedures incorporating the patient’s own tissues to recreate the breast mound (pedicled and free flaps, fat grafting). Others prefer a simple procedure, such as implant-based reconstruction, as they do not like having significantly long surgical procedures with associated donor-site morbidity. In these cases, subpectoral definitive silicone implant or temporary prosthesis is inserted in a pocket created below the pectoralis major muscle. In case of bilateral reconstructions as required after risk-reducing surgery, the difference in long-term outcomes with both options is not significant.102

When NAC sparing is not suitable, other surgical procedures are described.103 104 Carefully designed reduction patterns must be applied to retain a natural skin envelope (skin sparing mastectomies).105 Even more challenging are reconstructions performed following previous conservative surgery. In fact, mastectomy and reconstruction in irradiated breasts leads to higher postoperative complications due to impaired flap vascularity and healing after radiotherapy, the presence of previous scars and increased capsular contracture in case of implant-based procedures. The adverse effects of radiotherapy given both before and after implant-based reconstruction are well documented, and are significantly associated with a higher risk of reoperation.106 107 In the irradiated breast autologous reconstructions are the method of choice.

Complication rates after reconstruction vary considerably in the literature, from 8% to 64%.108 The most common complication is early wound necrosis/epidermolysis usually involving the NAC or skin edges, sometimes leading to implant loss.109 In addition, late complications may arise around cosmetic outcomes, almost always leading to additional surgery. In this respect, patients should be informed preoperatively that an optimal cosmetic reconstruction may not always be achieved with a single operation.


This international expert consensus on HLBC reviewed key questions regarding LBC caused by CDH1 germline variations, and the associated implications in asymptomatic mutant carriers.

A proposed workflow for the accurate clinical and genetic management of CDH1 carriers is presented in figure 4, approved by all members of this consensus.

The panel established that the latest IGCLC clinical criteria for CDH1 genetic screening are insufficient to identify/distinguish patients at risk of HLBC not associated with DGC. Therefore, novel criteria for genetic testing were suggested: (A) bilateral LBC with or without family history of LBC, with age at onset <50 years; and (B) unilateral LBC with family history of LBC, with age at onset <45 years.

Following genetic counselling, the identification of novel CDH1 germline variants should be classified in accordance with the established ACMG guidelines and discussed with the ClinGen Network. Currently, asymptomatic high-risk individuals carrying any pathogenic CDH1 alteration are not eligible for prophylactic mastectomy. Periodic breast examination, clinical surveillance and MRI are the preferred approaches. Gastric surveillance, as well as prophylactic gastrectomy, should be performed in accordance with the latest IGCLC guidelines (figure 4).

However, the panel established that in selected CDH1 mutant carriers and after multidisciplinary counselling, prophylactic surgery could be proposed. For example, contralateral prophylactic mastectomy can be an option for patients with diagnosis of LBC (also following previous ipsilateral breast surgery). Moreover, bilateral prophylactic mastectomy can be suggested in asymptomatic individuals with family history of LBC and well-documented CDH1 pathogenic alterations in a first-degree relative. In case of CDH1 mutations of unknown significance (so called VUS), clinical surveillance is still considered the best approach.


  1. 1.
  2. 2.
  3. 3.
  4. 4.
  5. 5.
  6. 6.
  7. 7.
  8. 8.
  9. 9.
  10. 10.
  11. 11.
  12. 12.
  13. 13.
  14. 14.
  15. 15.
  16. 16.
  17. 17.
  18. 18.
  19. 19.
  20. 20.
  21. 21.
  22. 22.
  23. 23.
  24. 24.
  25. 25.
  26. 26.
  27. 27.
  28. 28.
  29. 29.
  30. 30.
  31. 31.
  32. 32.
  33. 33.
  34. 34.
  35. 35.
  36. 36.
  37. 37.
  38. 38.
  39. 39.
  40. 40.
  41. 41.
  42. 42.
  43. 43.
  44. 44.
  45. 45.
  46. 46.
  47. 47.
  48. 48.
  49. 49.
  50. 50.
  51. 51.
  52. 52.
  53. 53.
  54. 54.
  55. 55.
  56. 56.
  57. 57.
  58. 58.
  59. 59.
  60. 60.
  61. 61.
  62. 62.
  63. 63.
  64. 64.
  65. 65.
  66. 66.
  67. 67.
  68. 68.
  69. 69.
  70. 70.
  71. 71.
  72. 72.
  73. 73.
  74. 74.
  75. 75.
  76. 76.
  77. 77.
  78. 78.
  79. 79.
  80. 80.
  81. 81.
  82. 82.
  83. 83.
  84. 84.
  85. 85.
  86. 86.
  87. 87.
  88. 88.
  89. 89.
  90. 90.
  91. 91.
  92. 92.
  93. 93.
  94. 94.
  95. 95.
  96. 96.
  97. 97.
  98. 98.
  99. 99.
  100. 100.
  101. 101.
  102. 102.
  103. 103.
  104. 104.
  105. 105.
  106. 106.
  107. 107.
  108. 108.
  109. 109.
  110. 110.
  111. 111.


  • BB and VG contributed equally.

  • Contributors GC and JF contributed equally. Concept and design: GC, BB, DM, VG. Supervisor board: PV, VG, MI, FR, FC, RS, CLV, GP. Iconography and graphic design: FS, JF, SM, EDC. Acquisition of data, analysis and interpretation of data, critical revision of the manuscript for important intellectual content, final approval of manuscript: all authors. Drafting of the manuscript: GC, RLG, JF, CLV, AT, AMDS, IF, EP, CT, KM, RK, FDL with input of all authors. Editorial assistance: MGV, EP.

  • Funding This manuscript was supported by the Italian Ministry of Health (Understanding how CDH1 germline mutations affect hereditary lobular breast cancer).

  • Competing interests None declared.

  • Patient consent Not required.

  • Provenance and peer review Commissioned; externally peer reviewed.

Request Permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.