Early detection of breast cancer-related lymphedema: Accuracy of indocyanine green lymphography compared with bioimpedance spectroscopy and subclinical lymphedema symptoms

Abstract

Introduction: The reported incidences of breast cancer-related lymphedema (LE) affecting the arms vary greatly. Reason for this variability includes different diagnostic techniques used across studies. In the current study, we compared the accuracy of indocyanine green lymphography (ICG_L) and bioimpedance spectroscopy (BIS) in detecting LE before presentation of clinical signs.Methods and Results: Patients with no initial detectable signs of clinical LE of their arms after axillary lymph node dissection or removal of >5 lymph nodes on sentinel lymph node biopsy were included. Subclinical LE was defined as BIS values outside the normal range [(>= 7 units (or >10 units)] or a 7-unit (or 10 unit) change between two measurements. We tracked ICG_L and BIS measurements for 133 potentially affected arms (n = 123). ICG_L detected signs of lymphatic flow disruption in 63 arms (47%). Based on the BIS value of 7 units, 60 arms (45%) had values outside the normal range. When using ICG_L-identified LE cases as true positives, BIS had a 54% accuracy (area under the curve [AUC] = 0.54) in detecting LE. Accuracy was 61% for subclinical LE symptoms when compared with ICG_L (AUC = 0.62). Both BIS and subclinical LE symptoms had <0.70 AUC-receiver characteristic operator curve, suggesting that BIS and development of subclinical LE symptoms are not adequate for identifying patients with subclinical LE.Conclusion: ICG_L is a reliable diagnostic tool for detecting early signs of lymphatic flow disruption in subclinical LE. Utilizing ICG_L to diagnose subclinical LE followed by a personalized treatment plan may provide patients the best chance of preventing disease progression.