Pathologists face daily challenges in diagnosing breast proliferative lesions based upon morphologic evidence as benign, borderline or malignant (in situ or invasive), a critical decision in determining the patient's management. Accurate identification of MEC has become one of the most helpful tools in making the diagnosis, as they are present and usually proliferate in benign lesions, become diminished in ADH and DCIS and are lost in invasive carcinomas. MEC are well known for their pleomorphism in size and shape, even in normal tissue, and their recognition can be very difficult in routine tissue sections. The development of immunohisto-chemistry has greatly improved the accuracy of their identification. Several MEC markers have been explored [1
]. The commonly used markers, including smooth muscle actin (SMA), smooth muscle myosin heavy chain (SM-MHC), calponin and p63, are highly sensitive but have varying specificities. For example, SMA stains scattered epithelial cells and reactive myofibro-blasts in the stroma [13
], invasive tumor cells may show focal calponin positivity [15
], and p63 may label cells of poorly differentiated carcinoma with squamous differentiation. To increase the specificity and sensitivity, a panel of antibodies is recommended.
D2-40 is a commercially available mouse monoclonal antibody directed against human podo-planin, a mucin-type transmemebrane protein originally reported in lymphatic endothelial cells [16
]. Besides staining lymphatic vessels, it has occasionally been reported to stain MEC of salivary gland and breast [9
]. We have also noted that when D2-40 is used to demonstrate lymphatic vessels, the MEC of normal and benign ducts and lobules of breast are stained. In this extended study, we demonstrated that MEC in benign breast tissues were positive for D2-40, while focally or diffusely diminished MEC were also identified by D2-40 in ADH and DCIS. As expected, no MEC were found in invasive carcinomas by D2-40. The pattern of D2-40 stain was highly correlated with those of calponin and p63. In comparison, the staining intensity of MEC was slightly less than that of calponin, and same was true for the background unspecific stain. The nuclear stain of p63 had the cleanest background and was the easiest to interpret. The study showed that D2-40 immunohistochemistry reliably labels MEC in a variety of breast tissues from benign to pre-malignant lesions, suggesting that it can be considered as an additional marker for MEC of breast.
LVI of breast cancer is an independent adverse prognosticator that is associated with increased regional and distant tumor recurrence and D2-40 immunohistochemisty has proven to enhance the detection of LVI. Braun et al [12
] reported that LVI was identified by D2-40 in 70 of 254 (28%) early breast cancer as compared to 40 tumors (16%) by routine HE staining, and D2-40 positivity was the strongest predictor for axillary lymph node metastasis in multivariate analysis. However, LVI detected by D2-40 may disqualify patients with early-stage breast cancer for accelerated partial breast irradiation [17
], and potential pitfalls in LVI interpretation by D2-40 immunohistochemistry have been recently discovered. Rabban and Chen [18
] found that the spindled, attenuated MEC surrounding solid DCIS resulted in a thin cytoplasmic D2-40 staining pattern that could be misinterpreted as LVI. They proposed that in these difficult cases p63 or SM-MHC should be used to confirm the presence of MEC and thereby exclude the diagnosis of LVI. Retraction artifact results in clear spaces around tumor cell nests and is frequently seen around invasive carcinoma or solid DCIS of breast. This artifact may be difficult to differentiate from LVI. Although a recent study suggested that retraction artifact associated with invasive carcinoma correlates with LVI and poor outcome of early breast carcinoma, the clinical significance of this artifact associated with DCIS is unknown [19
In our study, 5 cases of DCIS showed retraction artifact of different extent and the artifact space was surrounded by residual MEC and stroma with myofibroblasts. D2-40 stained the MEC attached to the periphery of the DCIS and the myofibroblasts in the stroma. When these cells were located along the artificial cavity, they mimicked lymphatic endothelial cells. However, the myofibroblasts were not labeled by p63 or calponin in our study. This potential mistake in interpretation can be avoided by the application of a panel of markers including another vascular marker, such as CD31, in difficult cases.
In summary, D2-40 immunohistochemistry reliably highlights the MEC of breast in a variety of breast lesions in a pattern similar to that of calponin and p63, and could be considered as an additional MEC marker. Should it be used, a panel with other MEC markers, such as calponin or p63 is recommended. Caution should be exercised when interpreting the staining of cells surrounding DCIS with retraction artifact. In difficult cases, additional markers for MEC and lymphatic vessels should be considered to avoid misinterpretation of LVI.