Abstract Archives of the RSNA, 2009
SSJ01-05
MRI Monitoring of Breast Cancer Response during Neoadjuvant Chemotherapy: Should Molecular Subtype Be Considered?
Scientific Papers
Presented on December 1, 2009
Presented as part of SSJ01: Breast Imaging (MR Image Interpretation)
Claudette Elisabeth Loo MD, Abstract Co-Author: Nothing to Disclose
Marieke E Straver MD, Abstract Co-Author: Nothing to Disclose
Jelle Wesseling, Abstract Co-Author: Nothing to Disclose
Sjoerd Rodenhuis MEd, Abstract Co-Author: Nothing to Disclose
Sara M. Muller PhD, Abstract Co-Author: Nothing to Disclose
Kenneth G.A. Gilhuijs PhD, Presenter: Nothing to Disclose
Marie-Jeanne Vrancken Peeters, Abstract Co-Author: Nothing to Disclose
00030490-DMT et al, Abstract Co-Author: Nothing to Disclose
To evaluate differences in predictive value of MRI features to monitor response of breast cancer during neoadjuvant chemotherapy (NAC) stratified to molecular subtype.
MR images of 188 women with proven breast cancer and scheduled for NAC were analyzed. Using core biopsies, tumors were stratified to molecular subtype: basal-type (estrogen-receptor negative), Her2+ (human epidermal growth factor receptor 2 positive), and luminal-type (estrogen-receptor positive).MR imaging was performed before and during NAC. Surgical specimens were examined to establish presence or absence of residual tumor. Interpretation of MRI included assessment of lesion morphology at baseline, changes in lesion morphology, size, and changes in contrast uptake kinetics. Multivariate binary logistic regression, cross-validation and ROC analysis were employed to establish significant associations.
Forty-seven women (25%) had a basal-type tumor, 38 (20%) a Her2+ tumor and 103 (55%) a luminal-type tumor. Residual tumor at pathology was found in 31 (66%), 23 (61%), and 96 (93%) specimens, respectively. Basal-type tumors were more often unifocal masses: 57% vs. 18% and 33% (p=0.001). Conversely, Her2+ tumors were more often multifocal masses: 53% vs. 32% and 30% (p=0.02). The RECIST criteria to monitor response were least effective for the luminal-type tumors ( (AUC=0.54), followed by the Her2+ (AUC=0.68) and the basal-types (AUC=0.78). Multivariate analysis showed distinct differences in MRI features predictive of residual tumor; Basal-type tumors: reduction of tumor size in combination with the pattern of reduction (AUC=0.86). Her2+ types: lesion morphology at baseline in combination with the pattern of reduction (AUC=0.85). Luminal types: no significant associations were found.
Distinct differences exist between MRI features predictive of tumor response and molecular subtype of the tumor. When molecular subtype is taken into consideration, changes in lesion morphology may provide useful complementary information to the RECIST criteria especially for Her2+ and basal-type tumors.
Knowledge of the molecular subtype of breast tumors provides useful complementary information to current MRI criteria for response-monitoring of breast-cancer to neoadjuvant chemotherapy.
Loo, C,
Straver, M,
Wesseling, J,
Rodenhuis, S,
Muller, S,
Gilhuijs, K,
Vrancken Peeters, M,
et al, 0,
MRI Monitoring of Breast Cancer Response during Neoadjuvant Chemotherapy: Should Molecular Subtype Be Considered?. Radiological Society of North America 2009 Scientific Assembly and Annual Meeting, November 29 - December 4, 2009 ,Chicago IL.
http://archive.rsna.org/2009/8005999.html