Rouzier R et al. (2005) Breast cancer molecular subtypes respond differently to preoperative chemotherapy. Clin Cancer Res 11: 5678–5685

Response to treatment can vary considerably among histologically similar breast cancers. Cancer subtypes have been identified and broadly classified into four categories on the basis of their gene expression profiles. Rouzier et al. examined fine-needle aspiration samples from 82 women with breast cancer, who subsequently underwent 24 weeks of preoperative paclitaxel and doxorubicin-containing chemotherapy. A previously identified 'breast intrinsic' set of 534 genes was used for expression profiling. Tumors were classified into four molecular classes: luminal (n = 30), normal-like (n = 10) basal-like (n = 22) and erbB2+ (n = 20) subtypes. These subgroups had different sensitivities to preoperative chemotherapy, with basal-like and erbB2+ subgroups showing the highest rates of pathological complete response (pCR; both 45%; 95% CI 23–68%) and luminal and normal-like subgroups showing the lowest rates (luminal 6%, 95% CI 1–21%; normal-like, no pCR, 95% CI 0–31%).

The pCR could be predicted from this subtype classification with equivalent accuracy to that obtained with conventional clinicopathological parameters. Logistic regression model analysis revealed that the molecular subgroups did not improve prediction accuracy compared with conventional parameters. The authors comment that this was because basal-like tumors were almost all estrogen-receptor negative, and erbB2+ and basal-like tumors were predominantly high grade—characteristics associated with a higher pCR.

Importantly, the genes that were associated with pCR in the basal-like and erbB2+ groups did not overlap. This suggests that the mechanisms of sensitivity to chemotherapy vary between these tumor subtypes.