“After 100 years, we need to do something” is how Dr Sunil Lakhani summed up a session on lobular breast cancer at Breast Cancer Trials’ recent scientific meeting (Brisbane, July 2026).
As a pathologist, Dr Lakhani explained the challenges in classifying variants of invasive lobular breast cancer (ILC) and lobular carcinoma in situ (LCIS). ILC accounts for 10 to 15% of breast cancer cases, but is there has been much less research on it than on the more common ductal type. (You can read more about ILC here.)
Most cases of ILC are hormone receptor positive and HER2 negative, with low Ki67 (slow rates of cancer cell division), and generally ILC is considered to have a good prognosis, but this might be a misleading belief.
Its morphology (shape and form) and clinical behaviour is very different from ductal breast cancer. ILC cells grow in single file, making the cancer hard to detect. Its growth can be indolent and persistent (slow-growing and slow-spreading), sometimes growing for years without obvious symptoms. Presentation is often late and surgical excision can be incomplete. Response to chemotherapy can be poor. Like ductal breast cancer, ILC can metastasise to bones, liver, lungs, brain and lymph nodes, but it can also spread to the gastrointestinal tract, the peritoneum, reproductive organs, leptomeninges (tissues covering the brain and spinal cord) and the skin and soft tissues. Some of these sites can be hard to treat.
LCIS is definitely a pre-cursor to ILC. LCIS is usually multifocal and often bilateral (both breasts). It is often diagnosed in women aged 40-50 years. Lobular breast cancer cells usually have mutations in the E-cadherin (CDH1) gene (60 to 80% of ILC cases have CDH1 mutations). E-cadherin’s normal function is to stick cells together but when it is mutated, cells can grow in abnormal ways. Samples of LCIS and ILC cells taken from single patients have been found to have identical mutations, indicating that LCIS is a true pre-cursor to ILC.
Unfortunately, LCIS is often referred to as a ‘benign condition’ or a ‘risk factor’ that simply increases the chance of getting breast cancer, rather than a clear pre-cursor to ILC. In contrast, ductal carcinoma in situ (DCIS) is considered a ‘Grade 0’ cancer and a definite pre-cursor to invasive cancer.
Dr Lakhani explained that this unhelpful belief has probably led to low or no funding for research on LCIS and ILC. (You can read more about lobular breast cancer and current NZ research on ILC here.)
Pathologists can see a variety of cell types and growth patterns in samples of lobular breast tissue, but classifying these into defined variants that could guide treatment decisions remains problematic.
Receptor staining shows that ILC can sometimes be HER2 positive or triple negative and “we treat these the same as ductal HER2 positive or triple negative breast cancers, even though we know that metastasis is not the same”, said Dr Lakhani.
To address these problems the European Lobular Breast Cancer Consortium was set up in 2018 to advance research, diagnosis and treatment for ILC. It works closely with patient advocacy organisations like the Lobular Breast Cancer Alliance and European Lobular Breast Advocates.
Dr Amy McCart Reed then spoke about targeting the molecular underpinning of ILC. By understanding more about the molecular signatures of ILC tumours, researchers are hoping to identify proteins that could be targeted by specific drugs to stop cancer cell growth and treat ILC.
She explained that in 30% of cases ILC cannot be seen in a mammogram and ILC is usually Grade 2 when found. There is a high rate of mastectomy for ILC and the prognosis can be poor. There is a clear need for techniques to distinguish different subtypes of ILC so that we can identify those at high risk of progression and poor outcomes and those at lower risk. A ‘prognostic signature’ is needed.
LobSig4 is a 14-gene panel developed from a 194-gene expression signature for ILC tumour tissues. The LobSig4 test can stratify cases into ‘high-risk’ and ‘low-risk’ categories for predicting likelihood of recurrence. LobSig4 has been designed specifically for lobular cancer, unlike the MammaPrint or Oncotype DX tests which were primarily built around ductal breast cancer. LobSig4 and its two-gene version LobSig2 still require extensive validation in large international cohorts, including women from different countries and of different ages, before it can be routinely implemented in standard clinical practice.
More research is also needed to identify ‘targetable’ proteins for ILC.
When lobular cancer cells mutate and lose the E-cadherin protein, they become completely dependent on the ROS-1 tyrosine kinase pathway to survive and proliferate. ROSALINE and REPLOT are clinical trials that looked at ROS-I inhibitors entrectinib and repotrectinib, but results so far have been disappointing, with REPLOT being stopped because of toxic side effects.
Experiments with mouse models, in which tumour cells from a patient are injected into mice, have shown that the LOXL1 gene encoding a LOX enzyme is important for lobular cancer cells’ survival. This suggests that LOX inhibitors might be useful therapies for ILC. PXS-5505 (amsulostat) is a LOX inhibitor being investigated for another disorder, so it could theoretically be used in a trial with ILC. However, more needs to be known about LOX expression in ILC tumours first.
Other research is looking at MDC1 (Mediator of DNA Damage Checkpoint Protein-1), a protein that normally promotes homologous recombination and repair of DNA breaks, but can become dysfunctional in breast cancers. This loss of function seems to be different in lobular and ductal cancers, suggesting different treatment strategies to target this might be useful. However, more comprehensive ILC specimen collections are needed before progress can be made in identifying and testing new therapies for ILC.
There is clearly a huge need for more research if we are to see better diagnosis and treatment of LCIS and ILC in the future.
12 September 2026
