Currently, when a woman is diagnosed with the earliest stage of breast cancer – called ductal carcinoma in situ – there is no way of knowing if it is going to be aggressive or not. Because of this, to be safe, most patients undergo aggressive treatment. Now, a new method that combines math with fluorescent microscopy can predict if this form of breast cancer is going to be aggressive or not.
The new method, called biomarker ratio imaging microscopy (BRIM), is the work of researchers at the University of Michigan, who describe how they used it to identify aggressive forms of ductal carcinoma in situ (DCIS) in a study published in the journal Scientific Reports.
One of the researchers, Howard R. Petty, professor of ophthalmology and visual sciences and of microbiology and immunology, explains how a patient with DCIS typically undergoes treatment as if the disease is invasive. This is easy to understand, he notes, and adds:
“When women hear breast cancer, they’re petrified. And physicians are keenly concerned about outcomes as well. But, DCIS is not the same disease for everyone. If we can identify potentially non-aggressive lesions, perhaps those women don’t need aggressive treatment.”
BRIM combines traditional microscope techniques that pathologists use to examine tissue with mathematical analysis. The technique compares levels of different biomarkers, which can be seen as different fluorescent colors in stained tissue under a microscope.
Ductal carcinoma in situ (DCIS) is a noninvasive condition where abnormal cells are found in the milk ducts of the breast. In situ means the abnormal cells have not spread to other tissues in the breast – that is, they are localized within the milk duct.
In some cases, DCIS may become aggressive and invade surrounding tissue, but currently, pathologists have no way of discerning which lesions could become invasive.
Long-term studies of women whose DCIS lesions were untreated because they were originally misclassified as benign found that 20-53 percent of them were diagnosed with an invasive breast cancer in the 10 years following original diagnosis.
DCIS can appear as a mass than can be felt, but it is most often detected from a mammogram, where it shows as tiny white dots signifying deposits of calcium. The deposits themselves are harmless, but they could indicate the presence of in situ or invasive cancer.
Because there is currently no certain way of knowing whether a DCIS lesion will lead to aggressive breast cancer, surgery and sometimes radiation and/or hormonal therapy is the usual treatment following a diagnosis.
DCIS accounts for over 80 percent of in situ breast carcinoma diagnoses, of which there were an estimated 60,290 cases in the United States in 2015. The other 20 percent or of cases are lobular carcinoma in situ (LCIS), which is generally not thought to be a precursor of invasive cancer.
For their study, the researchers looked at biopsy tissue samples from 23 patients with DCIS. They used fluorescent imaging – where the tissue samples are stained – to identify key biomarkers. Each biomarker was stained a different color.
They then entered the images of the stained tissue samples into a computer that calculated the levels of different biomarkers in each pixel.
In cancer, some biomarkers are present in high levels while others are less prolific. BRIM uses the ratio of these different levels to form an image of improved contrast.
The researchers used
The BRIM method found 22 percent of the samples had low ratios of cancer versus non-cancer biomarkers, suggesting those lesions were very slow-growing and non-aggressive.
The researchers note that an advantage of BRIM is that it uses several biomarkers rather than relying on only one. They decided which biomarkers to use after an extensive literature search.
“This approach is going to be a new and powerful one. It works because we’re looking at it mathematically.”
Prof. Howard R. Petty
As well as helping to avoid unnecessary treatment, BRIM could help with other breast cancer treatment decisions, say the researchers. Also, as more biomarkers for cancer are researched, it is likely that the method could be used with other types of cancer.
They are already planning a large study to look at links between BRIM scores and patient outcomes in past cases.