We appreciate the interest, thought, and detailed inspection of our methodology by Dr. Reis-Filho et al. They raise an interesting theory for our results. We were not aware of the cross-reactivity of the E-cadherin antibody utilized in our study, and had not considered the possibility that the reactivity we observed was due to P-cadherin. After receiving their correspondence, we immediately stained serial sections of the lesional blocks using the E-cadherin antibody clone they mentioned (clone HECD-1; Zymed Laboratory, Inc., San Francisco, CA). The area of interest was present on the deeper section slide in eight of the nine cases. There was similar or identical E-cadherin staining of the lobular carcinoma in situ (LCIS) cells with the HECD-1 clone as we initially observed with the 4A2C7 clone in all eight cases in our study. This finding supports our original theory that the LCIS membrane reactivity we observed was due to the presence of E-cadherin rather than P-cadherin. In addition, as Dr. Reis-Filho et al. note, P-cadherin reactivity is associated with high nuclear grade ductal carcinoma in situ with unfavorable biologic features. The lesions we studied were of low nuclear grade with low proliferation rates. We completely agree with their statement that further study is required before final conclusions can be drawn. We currently are evaluating the E-cadherin gene status of these and other lesions with the single-strand conformational polymorphism technique.

We do not view the results of our study as surprising or unique. Our study is only one of an increasing number of articles that have demonstrated a close association between low grade in situ and invasive ductal lesions and LCIS.1–8 Similar to our knowledge that the dyshesive growth pattern of some invasive carcinomas can result from the dysfunction or absence of several molecules or genetic mutations, we believe that LCIS finally will be understood to be a morphologic pattern that results from several different groups of molecular events rather than a single morphologic-genetic mutational lesion.

Lastly, we want to offer an opinion through which the results of our study should be interpreted: morphologic classification is the foundation and basis on which patients are treated. Ancillary studies, including immunohistochemistry and the numerous molecular techniques, afford the opportunity to acquire knowledge and an understanding of the interworkings that lead to the characteristic morphologic features of a lesion.9, 10 We did not intend for our study to be viewed as suggesting that ancillary molecular-related studies should replace morphology.


  1. Top of page

Neal S. Goldstein M.D.*, * Department of Anatomic Pathology, William Beaumont Hospital, Royal Oak, Michigan