Rational Use of Imaging to Stage Breast Cancer: Evidences for a Selective Approach

INTRODUCTION: Staging investigations at diagnosis are customary to accurately assign a clinical stage before therapy. The practice of routine imaging in patients asymptomatic for metastasis is not recommended but widely adopted. This study was done to reexamine the basis behind guideline recommendations and to identify the factors predictive of asymptomatic metastasis. METHODS: Oncology records of 200 breast cancer patients in clinical Stages I-III at diagnosis were prospectively reviewed. Baseline demographic information, tumor characteristics, and pathological data including molecular typing were collected. The prevalence of metastasis deduced and accuracy of bone scan, chest X-ray (CXR), liver ultrasound, and computed tomography (CT) chest analyzed. Patient and tumor characteristics predictive of asymptomatic metastasis tested for significance using appropriate statistical tests. RESULTS: The prevalence of asymptomatic metastasis was 13.5%. Bone lesions (8%) were the most common metastatic site followed by lungs (7%) and liver (1%). Sensitivity, specificity, positive- and negative-predictive values of bone scans and CT chest were 100%, 97%, 74%, 100%, and 92%, 99%, 87, 3%, 99.4%, respectively. The above values for ultrasound abdomen and CXRs were 100%, 99%, 93%, 100% and 21%, 94%, 20%, 94%, respectively. Tumor size (P = 0.001), tumor Stage T1/T2 versus T3/T4 (P = 0.0002), nodal stages N0/N1 versus N2/N3 (P = 0.001), high histological Grade G I versus GII/GIII (P = 0.0001) and molecular types were strongly predictive of metastatic disease. CONCLUSION: The routine use of imaging to detect distant metastasis in asymptomatic patients is not recommended in newly diagnosed breast cancer. A selective approach may be adopted in individuals with tumor more than 5 cm, advanced nodal disease, higher histological grade, and aggressive molecular types.