Immunohistochemistry for Diagnosing Breast Lesions
Immunohistochemistry for Diagnosing Breast Lesions
A myoepithelial layer is present around normal ducts and lobules, benign lesions, including sclerosing adenosis and radial scars, and ductal carcinoma in situ (DCIS), but not around invasive carcinoma. Sometimes the myoepithelial layer can be difficult to identify on H&E sections. Immunohistochemistry for myoepithelial cells can thus be of value in distinguishing invasive carcinoma from sclerosing lesions and DCIS (figures 1 and 2). It is important to be aware that there can be reduced staining around DCIS or sclerosing lesions compared with adjacent normal breast tissue.
(Enlarge Image)
Figure 1.
(A) H&E section shows groups of epithelial cells with architecture suspicious of invasive carcinoma. (B) Immunohistochemistry for smooth muscle actin shows staining around the normal duct in the centre and vessels and a thinner layer around some of the groups of epithelial cells. (C) Smooth muscle myosin heavy chain and (D) p63 show staining around the normal duct, but not around the other epithelial groups consistent with invasive carcinoma.
(Enlarge Image)
Figure 2.
Sclerosing lesion with lobular neoplasia. (A) H&E section shows a sclerosing lesion with a monotonous proliferation on the left. (B) Smooth muscle actin and (C) p63 show staining around all the acini, but there is attenuation around the spaces involved by lobular neoplasia. (D) E-cadherin is negative around the monotonous proliferation on the left consistent with lobular neoplasia and present in the normal epithelium on the right.
Myoepithelial markers are also useful in papillary lesions. Benign papillomas show a myoepithelial layer around the edge of the ducts and within the papillae. A benign papilloma involved by DCIS or atypical ductal hyperplasia shows a similar pattern, although there may be some attenuation of staining (figure 3). Papillary carcinoma in situ shows a layer around the edge of the ducts, but not within the lesion. An important pitfall is staining of blood vessels and occasionally myofibroblasts, which can be mistaken for myoepithelial cells (figure 4). Encysted or encapsulated papillary carcinoma typically shows no myoepithelial layer either at the edge or within the lesion. The distinction from invasive carcinoma must therefore be based on morphology. Basal cytokeratins play a complementary role in distinguishing epithelial hyperplasia of usual type from atypical ductal hyperplasia and DCIS within papillary lesions (see figure 3 and below).
(Enlarge Image)
Figure 3.
Papilloma with ductal carcinoma in situ. The section shows a papillary lesion with a monotonous epithelial proliferation. The benign epithelial component of the papilloma is not shown in the figure. A myoepithelial layer is shown both within the lesion and around the edge with (A) smooth muscle actin and (B) cytokeratin 14. The epithelium is uniformly negative with cytokeratin 14 consistent with a clonal proliferation.
(Enlarge Image)
Figure 4.
Papillary carcinoma in situ. Section shows a papillary lesion with a monotonous epithelial proliferation. (A) Smooth muscle actin shows staining in the fibrovascular cores. Some of the staining is in vessels, but there is not a continuous layer at the edge of the fibrovascular cores adjacent to the epithelium. (B) p63 shows no staining consistent with an absent myoepithelial layer.
Myoepithelial markers and luminal cytokeratins are useful for delineating the two components of adenomyoepitheliomas.
Smooth muscle actin is a robust myoepithelial marker and is often positive even in suboptimally fixed or infarcted tissue. Thus if there is no staining in well fixed tissue, this is good evidence that a myoepithelial layer is absent. The major weakness is lack of specificity, in particular the staining of myofibroblasts and blood vessels.
Calponin shows good sensitivity for myoepithelial cells with less staining of myofibroblasts than smooth muscle actin. Smooth muscle myosin heavy chain is more specific than smooth muscle actin with less staining of myofibroblasts, although vessels stain. p63 shows good specificity with no staining of myofibroblasts or vessels. Unlike the other myoepithelial markers which are cytoplasmic, p63 is a nuclear marker. As a result interpretation of staining can be difficult in some sclerosing lesions, particularly if the layer is attenuated. Both smooth muscle myosin heavy chain and p63 more often show reduced staining around DCIS or sclerosing lesions compared with smooth muscle actin and are more sensitive to poor fixation.
Antibodies to basal cytokeratins such as CK14 and CK5/6 stain myoepithelial cells, but are not reliable markers as they are frequently expressed at low levels. Also staining of epithelial cells can hamper interpretation.
It is most effective to use a panel of antibodies. I routinely use a combination of the sensitive marker smooth muscle actin, and two more specific markers such as smooth muscle myosin heavy chain and p63. As with all immunohistochemistry, it is important to be sure that the technique has worked by looking at internal controls, and if necessary, external controls. This is particularly important if the myoepithelial layer is absent in the lesion of interest. An important pitfall is that some carcinomas express myoepithelial markers. This group of basal-like carcinomas includes metaplastic carcinomas (see spindle cell lesion section below) and salivary gland-like carcinomas (figure 5). Basement membrane markers such as collagen IV and laminin cannot be used for distinguishing invasive carcinoma from DCIS or sclerosing lesions as some invasive carcinomas are surrounded by a layer of basement membrane (figure 5). A potential diagnostic pitfall is microglandular adenosis, which has traditionally been regarded as benign, although there is recent evidence that it may be a non-obligate precursor of carcinoma. This lesion is composed of tubules or islands of cells that are not in a lobular architecture and lack a myoepithelial layer, although basement membrane is present. The typical immunophenotype is positive for S100, cytokeratin 8/18 and focally for epidermal growth factor receptor or basal cytokeratins and negative for oestrogen receptor, progesterone receptor and HER2.
(Enlarge Image)
Figure 5.
Salivary gland-like carcinoma. (A) Smooth muscle myosin heavy chain is positive in some of the cells at the edge of the islands. (B) p63 shows more extensive expression. (C) There is a layer of collagen IV around the tumour islands.
Myoepithelial Markers
A myoepithelial layer is present around normal ducts and lobules, benign lesions, including sclerosing adenosis and radial scars, and ductal carcinoma in situ (DCIS), but not around invasive carcinoma. Sometimes the myoepithelial layer can be difficult to identify on H&E sections. Immunohistochemistry for myoepithelial cells can thus be of value in distinguishing invasive carcinoma from sclerosing lesions and DCIS (figures 1 and 2). It is important to be aware that there can be reduced staining around DCIS or sclerosing lesions compared with adjacent normal breast tissue.
(Enlarge Image)
Figure 1.
(A) H&E section shows groups of epithelial cells with architecture suspicious of invasive carcinoma. (B) Immunohistochemistry for smooth muscle actin shows staining around the normal duct in the centre and vessels and a thinner layer around some of the groups of epithelial cells. (C) Smooth muscle myosin heavy chain and (D) p63 show staining around the normal duct, but not around the other epithelial groups consistent with invasive carcinoma.
(Enlarge Image)
Figure 2.
Sclerosing lesion with lobular neoplasia. (A) H&E section shows a sclerosing lesion with a monotonous proliferation on the left. (B) Smooth muscle actin and (C) p63 show staining around all the acini, but there is attenuation around the spaces involved by lobular neoplasia. (D) E-cadherin is negative around the monotonous proliferation on the left consistent with lobular neoplasia and present in the normal epithelium on the right.
Myoepithelial markers are also useful in papillary lesions. Benign papillomas show a myoepithelial layer around the edge of the ducts and within the papillae. A benign papilloma involved by DCIS or atypical ductal hyperplasia shows a similar pattern, although there may be some attenuation of staining (figure 3). Papillary carcinoma in situ shows a layer around the edge of the ducts, but not within the lesion. An important pitfall is staining of blood vessels and occasionally myofibroblasts, which can be mistaken for myoepithelial cells (figure 4). Encysted or encapsulated papillary carcinoma typically shows no myoepithelial layer either at the edge or within the lesion. The distinction from invasive carcinoma must therefore be based on morphology. Basal cytokeratins play a complementary role in distinguishing epithelial hyperplasia of usual type from atypical ductal hyperplasia and DCIS within papillary lesions (see figure 3 and below).
(Enlarge Image)
Figure 3.
Papilloma with ductal carcinoma in situ. The section shows a papillary lesion with a monotonous epithelial proliferation. The benign epithelial component of the papilloma is not shown in the figure. A myoepithelial layer is shown both within the lesion and around the edge with (A) smooth muscle actin and (B) cytokeratin 14. The epithelium is uniformly negative with cytokeratin 14 consistent with a clonal proliferation.
(Enlarge Image)
Figure 4.
Papillary carcinoma in situ. Section shows a papillary lesion with a monotonous epithelial proliferation. (A) Smooth muscle actin shows staining in the fibrovascular cores. Some of the staining is in vessels, but there is not a continuous layer at the edge of the fibrovascular cores adjacent to the epithelium. (B) p63 shows no staining consistent with an absent myoepithelial layer.
Myoepithelial markers and luminal cytokeratins are useful for delineating the two components of adenomyoepitheliomas.
Smooth muscle actin is a robust myoepithelial marker and is often positive even in suboptimally fixed or infarcted tissue. Thus if there is no staining in well fixed tissue, this is good evidence that a myoepithelial layer is absent. The major weakness is lack of specificity, in particular the staining of myofibroblasts and blood vessels.
Calponin shows good sensitivity for myoepithelial cells with less staining of myofibroblasts than smooth muscle actin. Smooth muscle myosin heavy chain is more specific than smooth muscle actin with less staining of myofibroblasts, although vessels stain. p63 shows good specificity with no staining of myofibroblasts or vessels. Unlike the other myoepithelial markers which are cytoplasmic, p63 is a nuclear marker. As a result interpretation of staining can be difficult in some sclerosing lesions, particularly if the layer is attenuated. Both smooth muscle myosin heavy chain and p63 more often show reduced staining around DCIS or sclerosing lesions compared with smooth muscle actin and are more sensitive to poor fixation.
Antibodies to basal cytokeratins such as CK14 and CK5/6 stain myoepithelial cells, but are not reliable markers as they are frequently expressed at low levels. Also staining of epithelial cells can hamper interpretation.
It is most effective to use a panel of antibodies. I routinely use a combination of the sensitive marker smooth muscle actin, and two more specific markers such as smooth muscle myosin heavy chain and p63. As with all immunohistochemistry, it is important to be sure that the technique has worked by looking at internal controls, and if necessary, external controls. This is particularly important if the myoepithelial layer is absent in the lesion of interest. An important pitfall is that some carcinomas express myoepithelial markers. This group of basal-like carcinomas includes metaplastic carcinomas (see spindle cell lesion section below) and salivary gland-like carcinomas (figure 5). Basement membrane markers such as collagen IV and laminin cannot be used for distinguishing invasive carcinoma from DCIS or sclerosing lesions as some invasive carcinomas are surrounded by a layer of basement membrane (figure 5). A potential diagnostic pitfall is microglandular adenosis, which has traditionally been regarded as benign, although there is recent evidence that it may be a non-obligate precursor of carcinoma. This lesion is composed of tubules or islands of cells that are not in a lobular architecture and lack a myoepithelial layer, although basement membrane is present. The typical immunophenotype is positive for S100, cytokeratin 8/18 and focally for epidermal growth factor receptor or basal cytokeratins and negative for oestrogen receptor, progesterone receptor and HER2.
(Enlarge Image)
Figure 5.
Salivary gland-like carcinoma. (A) Smooth muscle myosin heavy chain is positive in some of the cells at the edge of the islands. (B) p63 shows more extensive expression. (C) There is a layer of collagen IV around the tumour islands.
Source...