[18F]FDG PET/CT volumetric biomarkers for non-invasive prediction of HER2 expression in breast cancer patients.

Abstract

Objectives: to investigate the capability of ¹⁸F-fluorodeoxyglucose positron emission tomography/computed tomography ([¹⁸F]-FDG PET/CT) derived volumetric parameters to predict human epidermal growth factor receptor 2 (HER2) status in breast cancer patients.

Methods: retrospective study enrolled 47 female patients with breast cancer. All patients had pretreatment [¹⁸F]-FDG PET/CT. Clinical data, pathology report and HER2 status were retrieved from medical records. In an attempt to assess the predictive value of the PET-derived metabolic parameters, Receiver operating characteristic (ROC) curve was constructed with area under curve analysis performed to detect best cutoff value of significant parameters for detection of HER2 positive.

Results: No statistically significant difference was noted among both groups (HER2 positive and negative) in respect to age, menopausal status, histology, grade, T-stage, N-stage, or antigen Kiel 67 (Ki-67) index. ROC curve successfully marked cutoff point ≥42.35 for total lesion glycolysis (TLG) and  12.75 for metabolic tumor value (MTV) that are capable to discriminate positive versus negative HER2 expression in breast cancer patients with area under curve (AUC) 0.728 and 0.723 and P-values 0.002 and 0.004 respectively. Such cutoff point was not deduced for standard uptake value (SUV) max. Primary tumor TLG cutoff correlated well with age where 77.8% of patients with TLG  42.35 were older than 45 years old compared to 22.2% of them who were younger than 45 years, P-value0.047. Also 70.3% of patients with TLG exceeds  42.35 had T3 and 4 primary tumors while 65% of those with TLG <42.35 their primary tumors were T1 and 2, P-value0.03. As regards Primary tumor MTV cutoff point, significant correlations were noted in respect to T-stage where 78.2% of the patients with primary tumor MTV  12.75 were T3 and 4, compared to 66.6% of those with primary tumor MTV <12.75 were T1 and 2, P-value0.011.

Conclusion: PET-derived volumetrics may serve as non-invasive predictors of biological processes represented here as HER2 expression in breast cancer patients.