Prognostic significance of preoperative 18F-FDG PET/CT for breast cancer subtypes

2016 
Abstract Adjuvant treatments for operable breast cancers are determined according to subtypes defined based on estrogen receptor (ER) and human epidermal growth factor receptor 2 (HER2) status. The ER+/HER2− subtype can be divided into luminal A and luminal B usually by Ki67 expression levels. Although tumor size, lymph node metastasis and tumor grade have been widely accepted in daily clinical practice, the identification of further prognostic indicators especially in the ER+/HER2− subtype is warranted. A total of 387 operated breast cancers for which maximum standardized uptake value (SUVmax) on the 18 F-fluorodeoxyglucose positron-emission tomography/computed tomography (FDG PET/CT) were available at baseline were retrospectively analyzed. The optimal cutoff value of SUVmax for relapse-free survival (RFS) was determined to be 3.585 by means of the receiver operating characteristics curve with an area under the curve of 0.6795 (95% CI: 0.5972 to 0.7618, p = 0.0006, sensitivity: 78.8%, specificity: 57.1%). The RFS of patients with SUVmax-high (n = 178) was significantly (p = 0.0003) worse compared with those with SUVmax-low (n = 209). This significant association was prominently recognized in the ER+/HER2− subtype. By multivariable analysis, SUVmax (hazard ratio: 3.83, 95% confidence interval: 1.28–11.51, p = 0.017), tumor size (4.22, 1.39–12.82, p = 0.011) and lymph node metastasis (4.44, 1.81–10.87, p = 0.0012) were significant and independent prognostic factors for RFS. The ER+/HER2− subtype demonstrated consistently worse RFS in the SUVmax-high patients both in the luminal A (p = 0.037) and luminal B (p = 0.047) subtypes. Combination of Ki67 and SUVmax appears to be useful for selecting patients who have inferior prognosis and need further adjuvant treatment of the ER+/HER2− subtype.
    • Correction
    • Source
    • Cite
    • Save
    • Machine Reading By IdeaReader
    22
    References
    21
    Citations
    NaN
    KQI
    []