P4-18-06: Relationship between Pathological Features, Her2 Protein Expression, and HER2 and CEP17 Copy Numbers in DCIS.

Abstract

Abstract Background Previous studies in which HER2 status in ductal carcinoma in situ (DCIS) was evaluated yielded conflicting results. The reported prevalences of HER2 overexpression and amplification vary remarkably. These discrepancies might partly be due to differences in assessment methods, i.e. FISH or immunohistochemistry (IHC) and usage of tissue microarray or whole mount slides. To further investigate this issue, we performed both FISH and IHC for HER2, evaluated HER2 and CEP17 copy numbers and correlated these data with histopathological findings. Methods 61 DCIS cases were studied, of which 55 pure DCIS and 6 DCIS with micro-invasion. Pathological features were evaluated and included: architecture, nuclear atypia, necrosis, calcifications, stromal inflammation, stromal morphology, extent of lesion, margin width, Van Nuys Prognostic Index and Pinder classification. Her2 IHC and HER2 dual probe FISH analysis were performed and scored according to ASCO/CAP guidelines. HER2/CEP17 ratio and HER2 and CEP17 copy numbers were separately analysed. IHC for estrogen and progesterone receptor (ER and PR) was also performed. Whole mount slides were used for all analyses. Results 15 cases (25%) were scored negative (1+), 10 cases (16%) equivocal (2+) and 36 cases (59%) positive (3+) using Her2 IHC. According to FISH analysis, 34 of 60 cases (57%) showed HER2 amplification; there was insufficient tissue for FISH analysis in one case. The amplification status of the DCIS lesions correlated with the IHC score (p<0,001). 30 of all 34 amplified cases were assigned a 3+ IHC score, and remarkably, all these cases showed HER2 clusters on FISH analysis (88%). Amplified lesions showed more frequently nuclear atypia grade 3 (p=0.0335), extensive comedonecrosis (p=0.002) and stromal inflammation (p=0.003). HER2 amplification correlated with the presence of micro-invasion (p=0.0218)%. There was no correlation with hormone receptor status or other pathological variables. In the amplified group, high-grade nuclear atypia was associated with a higher mean HER2 copy number (p=0.0026) and HER2/CEP17 ratio (p=0.0356), while this was not the case in the non-amplified group. CEP17 copy numbers did not correlate with nuclear atypia. Conclusions The correlation between HER2 amplification and adverse pathological features, including micro-invasion and the association in amplified DCIS between HER2 copy number and high-grade nuclear atypia, underscore that HER2 is a driver of DCIS aggressiveness and possibly of recurrence in the form of non-invasive cancer. However, the prevalence of HER2 overexpression, amplification and cluster formation was much higher than in invasive carcinoma, suggesting that HER2 might play a less important role in transition from DCIS to frankly invasive cancer. Further studies should evaluate non-invasive and invasive recurrence of resected DCIS separately. Citation Information: Cancer Res 2011;71(24 Suppl):Abstract nr P4-18-06.