New genetic test compares favorably to pathology in breast cancer diagnosis

If you go in for a mammogram and they see something that looks suspicious, odds are you are going to have to undergo a procedure called ductal lavage. Ductal lavage uses a fine needle to rinse the ducts of the breast with saline, and then has a pathologist look at the cells that come off to see if they are abnormal -- thus helping to detect breast cancer. One of the reasons that we do this is because mammography does not always detect cancers, and we would like to make sure that you do not have cancer. Ductal lavage is also less invasive than the alternative -- which is breast biopsy performed by a surgeon.

The big problem with lavage -- and to a degree biopsy -- is that you need a highly trained pathologist to look at the sample and see if it looks abnormal. I would like to emphasize how incredibly difficult this is. Pathologists go through years and years of training to learn how to do this, and, although most of them do an excellent job, no one is perfect. Sometimes, the cells in the ductal lavage don't look abnormal enough for the pathologist to give a good diagnosis. Sometimes a diagnosis of cancer is missed.

Just you can get a sense of what they are looking for here is a picture of a sample from ductal lavage:


The panel labelled C is normal cells. The panel labelled E is cells suggesting that they may be cancerous. This case is pretty clear cut. It isn't always this clear.

Eventually we would like to get to a point to where we can remove the human from the process of diagnosis. We would like to have lab tests for molecular markers for the cells that are more accurate than people. It is in this context that Fackler et al. published a molecular screen for ductal lavage cells that fairs not to badly against pathological diagnosis.

Detection of breast cancer is dependent on sensitive screening methods. Mammography is widely used but is falsely negative in 15% of women due to factors, such as breast density. This has led to a search for improved methods of imaging and sampling of breast tissue for cytologic examination. Cytology is the current gold standard for the identification of abnormalities typical of cellular transformation. However, recent findings have cast doubt on its effectiveness as a single discriminator of cancer cells. For instance, a recent study documented lack of reliability of cytology in cells collected by ductal lavage in women with biopsy-proven breast cancer. Here, similar to previous reports on cells derived through ductoscopy or fine needle aspiration, the sensitivity of cytology to detect cancer in ductal lavage was determined to be as low as 43%. Additional methods to identify tumor cells, such as those detecting molecular alterations, are clearly needed.

Multigene methylation of CpG islands is common in early breast cancer and leads to silencing of genes responsible for tumor suppression. Recently, we developed a highly sensitive method called quantitative multiplex methylation-specific PCR (QM-MSP) to quantitate cumulative gene promoter hypermethylation in multiple genes in samples where DNA is limiting, such as ductal lavage, nipple aspiration fluids, and fine needle aspirates. This method, using a five-gene panel consisting of RASSF1A, RAR-Beta, TWIST, HIN1, and Cyclin D2, evaluated percent methylation values in each gene and provided a clear distinction between normal and tumor tissues. (Citations have been removed.)

This test exploits an observation from genetics that DNA is methylated and that the degree of methylation in genes often relates to their level of expression. We know that certain genes are disregulated in cancer, so if you were to measure the level of methylation of certain genes you should theoretically be able to classify a cell as cancerous or normal.

The key issues with any screening test are sensitivity and specificity. Sensitivity reflects how likely the test is to detect the disease if it is there. Specificity reflects how likely it is to correctly diagnose a disease as not being present. Ideally, you want a test to do both well because you both don't want to miss cancer if it is there and don't want to subject someone to unnecessary surgery if it isn't. However, there is often a trade off between the two.

By these standards, this new test is hardly perfect, but neither is pathological diagnosis.

In the mastectomy group, a three-way comparison among duct histology, lavage cell cytology, and methylation showed the following. Cytology of ductal lavage fluid detected cancer in ducts proven by histology to have ductal carcinoma in situ or invasive cancer in 33% of the cases (7 of 21 ducts), with a specificity of 98.6% (72 of 73) and accuracy of 84.0% (79 of 94). Thus, cytology was found relatively insensitive but highly specific in the ductal lavage setting, confirming earlier findings. Although 3 of the 37 ductal lavage samples had insufficient cellular material for diagnosis by cytology, methylation results were informative. QM-MSP analysis was positive for methylation in ductal lavage with a sensitivity of 57.1% (12 of 21), specificity of 82.9% (63 of 76), and accuracy of 77.3% (75 of 97). (Emphasis mine. Citations have been removed.)

These results reveal that the new test is not as specific as specific as the pathologist, but it is more sensitive.

It would be premature to say that this test will completely supplant pathologists. Odds are it could be used to assist pathologists. But these types of technology are improving, and in the process they are improving our ability to remove human error from the process of diagnosis.

Definitely an area to keep watching.

More like this

I've written many times about how the relationship between the early detection of cancer and decreased mortality from cancer is not nearly as straightforward as the average person—even the average doctor—thinks, the first time being in the very first year of this blog's existence. Since then, the…
I've written more times than I can remember about the phenomenon of overdiagnosis and the phenomenon that is linked at the hip with it, overtreatment. Overdiagnosis is a problem that arises when large populations of asymptomatic, apparently healthy people are screened for a disease or a condition,…
After a lot of the not-so-Respectful Insolence⢠of the last couple of weeks, I've been meaning to get back to living up to the name of the overall mega-blog, namely ScienceBlogs. Meeting up with my fellow SB'ers over the weekend in New York Fortunately, last week a topic just so happened to pop…
A new study from Johns Hopkins shows that by sampling oral tissue from patients with a history of head and neck cancer doctors can predict with some accuracy the presence of a recurrence or of another primary tumor (presumably from the oropharynx). I found a couple of things about this report…

Jake, I have to take issue with your conclusions in this post. The problem is that women who have an abnormality on mammography don't get a ductal lavage; they get a biopsy. That means cutting out a hunk of tissue (lumpectomy) and having a pathologist look at it under a microscope, and it is the best single way (actually the only way) of getting a breast ca diagnosis. In fact it's the gold standard against which the tests you mentioned above are compared. My quick literature review (very quick, I grant you) suggests that lavage is really still an experimental procedure, and I've never heard of a clinician actually using it when there's a suspicion of breast ca. I'm not that far out of my residency (only 2 years) and I can tell you that this is the first time I've ever heard of ductal lavage.

I'm actually not sure where the authors get off saying that "Cytology is the current gold standard for the identification of abnormalities typical of cellular transformation." The current and ONLY gold standard is biopsy and pathologic examination unless they're trying to say something that I'm misunderstanding. It seems to me that this genetic test is pretty crappy. The last thing we want is a test that only picks up only 60% of breast cancers when they are present, and this figure will only get worse if it's actually used in a real world population. (The general rule is that as soon as you take the test out of the lab and start using it on real patients, the performance characteristics get worse.) The fact that it's better than lavage is irrelevant, 'cause lavage stinks too!

About 5% to 10% of breast cancer cases are reported to be hereditary. In these cases, the disease run in the families and directly results from mutated genes inherited from a parent. Genetic cancer testing has become a popular means of knowing whether or not you have cancer in your genes. But before you go for such a test, it is important to have a basic understanding about the test and its consequences.

First of all, a candidate must undergo intensive genetic counseling before taking genetic testing for breast cancer. This will be an educational session where the pros and cons of genetic cancer testing will be explained to the candidate.