| Home | E-Submission | Sitemap | Contact Us |  

The Korean Journal of Pathology 1999;33(6): 434-442.
Ductal Carcinoma In Situ of the Breast: Comparison of Histologic Classifications and Correlation with Histologic Grade of Coexisting Invasive Ductal Carcinoma.
Sung Ran Hong, Yee Jeong Kim, Yi Kyeong Chun, Hye Sun Kim, Hy Sook Kim
Department of Pathology, Samsung Cheil Hospital, Sungkyunkwan University School of Medicine, Seoul 100-380, Korea.
Recently developed new classifications (Holland, Van Nuys, modified Lagios) of ductal carcinoma in situ (DCIS) linked to outcome have emphasized the importance of nuclear morphology rather than architecture. We have evaluated these three classifications in ductal carcinomas composed of in situ and invasive carcinomas. The reproducibility of three classifications was assessed (n=49), and the histological grade of the DCIS was compared with the histologic differentiation (modified Bloom & Richardson method) and nuclear grade (modified Black method) of the coexisting invasive ductal carcinoma (n=45). According to Holland classification, the DCIS component was poorly differentiated in 51.0%, intermediately differentiated in 40.8%, and well differentiated in 8.2%. Using the Van Nuys classification, the DCIS component was group 3 (high grade with or without necrosis) in 44.9%, group 2 (non-high grade with necrosis) in 28.6%, and group 1 (non-high grade without necrosis) in 26.5%. According to the modified Lagios classification, the DCIS component was high-grade in 42.8%, intermediate-grade in 32.7%, and low-grade in 24.5%. The histologic grades of the three classifications revealed significant correlations between Holland and Van Nuys classification (p<0.0001) and between Holland and modified Lagios classification (p<0.0001), especially in poorly differentiated/group 3/high-grade DCIS. The reproducibility of classification of the DCIS was 71.4% in the Holland, 61.2% in the Van Nuys, and 55.1% in the modified Lagios classifications. The grade of the DCIS showed significant correlation with the grade of coexisting invasive ductal carcinoma (p<0.0001), especially in poorly differentiated/group 3/high-grade DCIS. In conclusion, DCIS grade, determined by the Holland, Van Nuys or modified Lagios classifications, is closely correlated with the histologic grade of the invasive ductal component in tumors composed of in situ and invasive ductal carcinoma, and may be a useful factor to estimate clinical behavior of DCIS. In our experience the Holland classification is recommended for DCIS classification due to its high reproducibility.
Key Words: Ductal carcinoma in situ; Reproducibility; Classification; Invasive ductal carcinoma