18F-FDG PET-CT imaging in the neoadjuvant setting for stages II-III breast cancer: association of loco­regional SUVmax with classical prognostic factors.

ABSTRACT

AIM: Classical prognostic and predictive factors serve to predict outcome and response to neoadjuvant treatment in stage II and III breast cancer. The purpose is to determine the relation between the SUVmax of the locoregional disease with these classical prognostic factors. METHODS: A prospective study including 43 stage II and III breast cancer patients was performed. In all the patients, two 18F-FDG PET-CT studies were performed before and after neoadjuvant chemotherapy. After this treatment, surgery and adjuvant treatment were carried out. To combine the information of the locoregional disease, the SUVmax of the lesion (tumor or abnormal lymphnodes) with the highest uptake was used. SUVmax and prognostic factors were studied with the Kruskal-Wallis non-parametric test and with the Mann-Whitney U. RESULTS: A statistically significant association between elevated SUVmax value and absence of estrogen receptors (ER) expression (16 vs.. 10; P<0.019) was found. Locorregional disease with positive HER2 phenotype had a statistically significant SUVmax value greater than Luminal A (estrogen and/or progesterone positive receptors with Ki67<15%) and B (estrogen and/or progesterone positive receptors with Ki67≥to 15%) (13.4 vs. 7.9 and 8.9; P<0.022 and P<0.024, respectively). Triple negative phenotype disease had higher SUVmax than Luminal A and B (15.4; P<0.030 and P<0.038). Positive correlation between the percentage of the Ki67 Proliferation Index and SUVmax (P<0.007) was demonstrated. High grade disease had a higher SUVmax than low grade (P<0.004). CONCLUSION: Locorregional SUVmax is associated to prognostic and predictive factors and reaffirms the utility of PET-CT as a tool in the common clinical practice.