Rosen's Breast Pathology, 4e

389

Ductal Carcinoma In Situ

FIG. 11.73.  Invasive ductal carcinoma with DCIS pattern. A: Extensive carcinoma with this appearance in the breast was interpreted as DCIS. Immunostains revealed absence of myoepithelial cells and basement membrane around some glandular structures, indicative of an invasive com- ponent. B,C: Metastatic carcinoma in ALNs duplicated the DCIS-like appearance of the primary invasive tumor. 5. Immunostains for basement membrane components, laminin and type IV collagen, are sometimes helpful. Absence of reactivity for both components indicates a strong likelihood of invasive carcinoma, especially if coupled with absence of myoepithelial cells. A ­reticulin stain may also be helpful in this setting, as well as whenever microglandular adenosis is a diagnostic consideration. 6. Reactivity for one or both basal lamina components in the absence of myoepithelial cells presents the most difficult diagnostic situation that requires assessment of the entire lesion, including multiple H&E levels if possible (see #2). The presence of laminin and type IV collagen favors a diagnosis of in situ carcinoma. How- ever, consideration must be given to the possibility that basal lamina may be formed at sites of invasion. With presently available routine diagnostic techniques, the distinction between basal lamina formed at sites of in- vasion and basement membranes in in situ carcinoma cannot be resolved with confidence in all cases. It is recommended that the term microinvasion be used for invasive lesions 1 mm or less in largest extent. This definition has been adopted by the TNM staging system with the rubric T1 mic to provide a descriptive identity for these unusually small invasive lesions that are otherwise

It is essential to use more than one immunostain, since reactivity is not equally intense with all reagents. 2. The absence of demonstrable immunoreactivity with an appropriate marker usually means that myoepithe- lial cells are not present, although they can be severely attenuated and difficult to recognize. Loss of the myo- epithelial cell layer occurs in some but not all DCIS and in certain types of benign (e.g., cystic apocrine lesions) and noninvasive neoplastic (e.g., some forms of papil- lary) processes. By itself, absence of myoepithelial cells is not indicative of invasive carcinoma, and the interpre- tation of this finding depends on the assessment of all histologic appearances of the lesion in the corresponding H&E section. 3. A new consecutive H&E section must be prepared whenever immunostains are done for suspected mi- croinvasion. This is necessary because the structure of the lesional tissue changes as additional slides are made. 4. Cytokeratin immunostains are essential for the evalua- tion of any focus suspected to be the site of microinva- sion. It is recommended that at least two different stains be used (e.g., CK7 and AE1/3) because of the variable reactivity of carcinoma cells. Cytokeratin immunostain- ing highlights the distribution of epithelial cells and dis- tinguishes epithelial cells from histiocytes.