FDG PET/CT and diffusion-weighted imaging for breast cancer: prognostic value of maximum standardized uptake values...

To correlate both primary lesion (18)F-fluorodeoxyglucose (FDG) maximum standardized uptake value (SUVmax) and diffusion-weighted imaging (DWI) apparent diffusion coefficient (ADC) with clinicopathological prognostic factors and compare the prognostic value of these indexes in breast cancer.

The study population consisted of 44 patients with 44 breast cancers visible on both preoperative FDG PET/CT and DWI images. The breast cancers included 9 ductal carcinoma in situ (DCIS) and 35 invasive ductal carcinomas (IDC). The relationships between both SUVmax and ADC and clinicopathological prognostic factors were evaluated by univariate and multivariate regression analysis and the degree of correlation was determined by Spearman's rank test. The patients were divided into a better prognosis group (n = 24) and a worse prognosis group (n = 20) based upon invasiveness (DCIS or IDC) and upon their prognostic group (good, moderate or poor) determined from the modified Nottingham prognostic index. Their prognostic values were examined by receiver operating characteristic analysis.

Both SUVmax and ADC were significantly associated (p<0.05) with histological grade (independently), nodal status and vascular invasion. Significant associations were also noted between SUVmax and tumour size (independently), oestrogen receptor status and human epidermal growth factor receptor-2 status, and between ADC and invasiveness. SUVmax and ADC were negatively correlated (ρ=-0.486, p = 0.001) and positively and negatively associated with increasing of histological grade, respectively. The threshold values for predicting a worse prognosis were ≥4.2 for SUVmax (with a sensitivity, specificity and accuracy of 80%, 75% and 77%, respectively) and ≤0.98 for ADC (with a sensitivity, specificity and accuracy of 90%, 67% and 77%, respectively).

SUVmax and ADC correlated with several of pathological prognostic factors and both indexes may have the same potential for predicting the prognosis of breast cancer.