Search the Health Library
Get the facts on diseases, conditions, tests and procedures.
I Want To...
I Want To...
Find Research Faculty
Enter the last name, specialty or keyword for your search below.
School of Medicine
Frequently Asked Questions - Pathology
1. Should we do frozen sections on breast biopsies?
The answer is almost always "NO!". Frozen sections used to be done to establish the diagnosis of invasive carcinoma, so that the surgeon could proceed to mastectomy while the patient was under anesthesia, and the pathologist could select fresh tissue for biochemical estrogen and progesterone receptor analysis. However, both of these reasons are now outdated. First, radiologically-guided needle core biopsy now allows us to establish the diagnosis of invasive carcinoma without the need for open biopsy. Options for therapy (i.e., lumpectomy plus radiation versus mastectomy) can therefore be evaluated before definitive surgery. Second, we now can test for estrogen and progesterone receptors on fixed tissue using immunohistochemistry, so fresh tissue is not needed for this purpose. Hence, the former benefits of frozen section are no longer relevant. Frozen sections do have significant disadvantages. First, frozen sections are difficult to interpret, and are more likely to be misinterpreted by a pathologist than are usual permanent slides. For example, benign inflammatory cells can be difficult to distinguish from infiltrating lobular carcinoma on frozen sections. Second, the process of freezing tissue introduces considerable artifacts in it that can make the lesion impossible to diagnose. For example, it is well known that frozen section artifact can make it impossible to distinguish a benign papilloma from a papillary carcinoma, because it is hard to appreciate the myoepithelial cells that accompany a papilloma and which are absent in a papillary carcinoma. Also, immunohistochemical stains may not always be easily interpreted in tissue that has been frozen. Finally, small lesions, such as are detected mammographically, may be depleted or even completely lost in the process of rapidly trimming the tissue in a frozen section cryostat. Such lesions are better handled carefully under standard processing (parenthetically, a similar argument can be made against performing frozen sections on sentinel lymph nodes. Hence, we generally do not perform frozen sections on breast specimens at Johns Hopkins.
2. What is the "S phase fraction (SPF)"?
This is basically a measurement of how rapidly the tumor cells are dividing or proliferating. S phase is the part of the cell cycle where DNA is synthesized; this occurs just before a cell divides into two. S phase fraction is measured by flow cytometry. It is easiest to perform on fresh tissue, but can be applied to formalin-fixed tissue blocks, though the results may be suboptimal. Another downside of this technique is that there are really no interlaboratory standards for what is a normal SPF; different laboratories establish their own ranges. At Johns Hopkins, we use the Ki-67 immunohistochemical stain to measure proliferation. We favor technique because it allows us to assess proliferation only of the tumor cells, since we see the tissue on the slide that stains for Ki-67.
3. What are the prognostic factors in breast carcinoma?
It is important to distinguish between a prognostic factor, which forecasts outcome independent of the therapy used, and a predictive factor, which forecasts response to a specific therapy (for example, Tamoxifen or Herceptin)
Established Prognostic Factors in Breast Cancer
1. Stage-how far the tumor has spread. - This is determined by three factors, abbreviated as TNM.
- Tumor size. This refers to size of the invasive carcinoma, not the in situ (non-invasive, confined to within the ducts and lobules) component or the desmoplastic reaction (scar). Smaller is better. For large tumors (>3cm), the gross size of the tumor is generally used. For tumors less than 3cm, the microscopic measurement is generally more accurate, since we can exclude the in situ component and desmoplastic reaction to the tumor.
- Lymph Nodes. This refers to whether or not there are metastases in the axillary lymph nodes (located under the arm) that the surgeon removed. It is better not to have metastases. If they are present, it is usually better if only a few nodes are involved and if the metastases are small.
- Metastases. This refers to spread outside of the breast to distant sites like the lung, liver, or bone. The prognosis is better when metastases are absent.
2. Tumor type
Most breast cancers are either typical ductal carcinomas or lobular carcinomas. However, there are a few more rare variant types of cancer that are associated with a better prognosis. These are Medullary carcinoma-These tumors feature a circumscribed rounded border, contain undifferentiated cells featuring high-grade nuclei and syncytial (sheet-like without well-defined cell borders) growth pattern, have scant fibrosis, and are associated with a dense infiltrate by lymphocytes and plasma cells. While this appearance is worrisome under the microscope (they would otherwise be scored as Elston 3 of 3), tumors with this appearance throughout have a paradoxically good prognosis Tubular carcinoma- These tumors feature tubule formation in over 90% of their area, and have bland, low-grade nuclei. Mitoses are rare Cribriform carcinoma-These tumors feature sieve-like collections of bland tumor cells. It often coexists with tubular carcinoma. Adenoid Cystic Carcinoma. Under the microscope, this tumor is identical to a very aggressive tumor of the same name that arises in the salivary gland. The tumor features a glandular (adenoid) component and a stromal/basement membrane (cylindromatous) component. Colloid (Mucinous) Carcinoma-This tumor is associated with extensive extracellular mucin production, usually occur in older women, and grow with a smooth pushing border. Over 90% of the tumor should be associated with extracellular mucin, and over 30% of the area of the tumor should be mucin, not tumor cells. These tumors also have have low-grade nuclei.
The pathologist assigns the cancer an Elston grade on a scale of 1 to 3, based upon how aggressive it looks under the microscope. Elston grade 1 is best ("well-differentiated"), while Elston grade 3 is worst ("poorly-differentiated").
The pathologist considers three factors in determining the grade. Each factor is given a point score on a scale of 1 to 3. The point score for each factor is then added to give the final score, as follows
How much of the tumor makes duct structures, like normal breast epithelium does?
- More than 75% (1 point)
- Between 10 and 75%? (2 points)
- Less than 10% (3 points)
Mitotic activity. How many mitotic figures (cell divisions) does the pathologist see in 10 high power (400X magnification) fields. The point score cutoffs depend upon the size of the high power field of the microscope that is used. In general,
- < 5 mitoses per 10 high power fields = 1 point
- 5-10 mitoses/10 high power fields = 2 points
- > 10 mitoses/10 high power fields = 3 points
Nuclear pleomorphism. Do the tumor nuclei look uniform similar to the normal breast duct epithelial cells, or are they larger, more irregular, and darker (hyperchromatic)?.
- Uniform nuclei similar to normal breast epithelium = 1 point
- Moderate nuclear variation = 2 points
- Marked nuclear variation, hyperchromasia, irregular large nucleoli = 3 points
These are added to yield a Final Elston Grade:
- Final score 3-5=Elston grade 1
- Final score 6-7=Elston grade 2
- Final score 8-9=Elston grade 3
4.Estrogen and Progesterone receptor (ER and PR)
Normal breast epithelium usually expresses ER and PR, and these mediate the normal process of cell division of these cells. We measure both of these receptors because the synthesis of progesterone receptor (PR) requires intact estrogen receptor (ER) signaling. Tumors that express these can be thought of as closer to normal breast than those that don't, and hence are associated with a better prognosis. Hence, positive ER and PR are favorable prognostic factors. About two thirds of breast cancers will express these receptors. The receptors are key predictive factors for predicting response to Tamoxifen therapy.
This oncogene (a gene which can promote cancer) is amplified (more copies of its DNA are present) and overexpressed (more of its protein is made) in 25-30% of human breast carcinomas. Amplification is a poor prognostic factor, though a weak one. Patients with positive lymph nodes have a worse prognosis when her2neu is amplified. Amplification of her2neu does predict response to therapy with Herceptin, the humanized anti-her2neu antibody that is now being used for patients with metastatic breast carcinoma. Her2neu is tested for in two ways. The most common way is by immunohistochemistry for Her2neu protein overexpression, using an antibody that binds to the protein. Many different antibodies can be used, and the results are not always the same with different antibodies, so it is important to know the result and what antibody was used in the test. Scores may be graded as strong positive, weak positive, and negative. Some laboratories test all cases that have proven to be weak positives by immunohistochemistry with the other method, FISH (Fluorescence In Situ Hybridization). This is a genetic test done in specialized laboratories in which one uses a probe to the Her2neu gene to determine how many copies of the gene are present in each cell. The FISH test determines if there is DNA amplification of the Her2neu gene (which almost always results in protein overexpression).
6. Ki-67 index
This is a measure of cell proliferation or growth rate. This antibody labels all cells actively participating in cell division. This includes the G1 (pre-DNA synthesis), S phase (DNA synthesis), G2 (post-synthesis), and M (mitosis) phases of the human cell cycle. It does not label cells in the resting (G0) phase of the cycle. This immunohistochemical stain correlates with the number of mitoses we count and incorporate into the grade, but is more quantitative; i.e., we can give a percentage of positive cells.
Other Potential Prognostic Factors:
Numerous other factors have been studied for their ability to predict outcome of patients with breast cancer. These include p53 gene status, DNA ploidy analysis, tumor angiogenesis, bcl2, cathepsin D, and epidermal growth factor receptor. To date, none of these new markers adds information to that which is already available through the markers described above, so they are not routinely performed at Johns Hopkins.