NEW YORK (GenomeWeb) – Researchers led by the University of North Carolina at Chapel Hill's Charles Perou have profiled more than 800 breast tumors to paint a molecular portrait of invasive lobular breast cancer.
Perou and his colleagues reported today in Cell that the invasive lobular breast carcinomas (ILCs) in their sample harbored the hallmark loss of E-cadherin as well as mutations in FOXA1, GATA3, PTEN, and more.
Some mutations the researchers found appeared to underscore how estrogen receptor activity differs between ILC and invasive ductal carcinoma (IDC), as well as define subtypes of invasive lobular breast carcinoma.
"This study identified multiple genomic alterations that discriminate between ILC and IDC demonstrating at the molecular level that ILC is a distinct breast cancer subtype and providing new insight into ILC tumor biology and therapeutic options," Perou and his colleagues wrote in their paper.
The researchers profiled 817 breast tumor samples — 127 ILC, 490 IDC, 88 mixed IDC/ILC, and 112 samples with other histology — through whole-exome sequencing, RNA sequencing, miRNA sequencing, SNP arrays, and DNA methylation arrays. A subset of 633 samples was also analyzed using reverse-phase protein array.
This, they noted, is a deeper dive into ILC than what The Cancer Genome Atlas was able to do with its 36 samples.
Within the ILC samples, Perou and his colleagues found more than 8,000 coding mutations. A number of mutations, they added, had different incidence rates in ILC, as compared to IDC.
ILC cases, for instance, were significantly enriched for CDH1 mutations — loss of the epithelial-specific cell-cell adhesion molecular E-cadherin (CDH1) is a hallmark of ILC. In nearly all the ILC cases, the researchers reported observing CDH1 loss at the DNA, mRNA, and protein levels. This loss, they added, is thought to lead to the tumor's characteristic discohesive morphology.
However, Perou and his colleagues couldn't reproduce previous findings that indicated high DNA methylation at the CDH1 promoter might be an alternative way to downregulate CDH1 in ILC.
ILC and IDC also differed in their incidence of FOXA1 mutations. FOXA1, the researchers noted, is a key ER transcriptional modulator that coordinates ER DNA binding as well as mediates long-range DNA interactions.
They uncovered 33 FOXA1 mutations in 30 of the 817 tumors they analyzed. The mutations, they noted, tended to cluster in the fork-head DNA binding (FK) and C terminus transactivation domains, but all of the ILC FOXA1 mutations were in the FK domain. Within that domain, the mutations further huddled in the W2 wing loop.
However, these changes are unlikely to affect FOXA1 DNA binding, as these FOXA1 mutations were positively associated with FOXA1 mRNA expression.
Instead, the researchers suggested that the FOXA1 mutations may instead activate alternative mechanisms that affect ER transcriptional programs.
At the same time, the researchers found that ILC samples had fewer mutations in GATA3 — another ER modulator — than IDC samples, suggesting to them that ILC and IDC have different ER modulator requirements.
PTEN mutations could also distinguish luminal A ILC and luminal A IDC tumors, as ILC tumors had lower PTEN protein expression and increased Akt signaling. The PI3K/Akt pathway is among the most altered in cancer as it confers enhanced growth and survival abilities on tumor cells, the researchers noted.
From their mRNA-seq expression data, the researchers uncovered three ILC subtypes — reactive-like, immune-related, and proliferative — and developed signatures that could identify these subtypes.
The immune-related ILC tumors, for instance, exhibited high expression levels of interleukins, chemokine receptors and ligands, the major histocompatibility complex, and tumor necrosis factors, among others.
These different subtypes were also linked to different clinical outcomes. Patients with reactive-like ILC had better disease-specific and overall surgical outcomes, compared to proliferative ILC patients.
Perou and his colleagues also found that the mixed ILC/IDC tumors largely had molecular features that resembled one group or the other and don't represent a hybrid of the two, a finding they said could help inform treatment decisions.
"This multi-platform study identified numerous molecular features discriminating between breast ILC and IDC, demonstrating different pathways underlying their pathogenesis, defining new ILC subtypes with different clinical outcomes, and pointing to previously unrecognized therapeutic possibilities," the researchers added.