Abstract
AIMS: This study aimed to evaluate the diagnostic accuracy of the apparent diffusion coefficient (ADC) derived from diffusion weighted imaging (DWI) for detecting lymph node metastasis in breast cancer. METHODS: A systematic review and meta-analysis was conducted following the PRISMA-DTA guidelines. PubMed, Web of Science, PROQUEST, and EMBASE were searched. ADC values for suspected lymph nodes in breast cancer patients were assessed, with histopathology as the reference test. Data were analyzed using a bivariate meta-analytical model with Meta-DiSc 2.0 and STATA 17. Meta-regression and subgroup analysis were conducted to assess heterogeneity. RESULTS: Twenty-six studies encompassing 2828 participants were included. The pooled sensitivity and specificity of ADC value for detecting lymph node metastasis were 88.6% (CI = 82.4%-92.8%) and 83.6% (CI = 78.0%-88.1%), respectively. The area under the HSROC curve was 0.92 (CI = 0.89-0.94), indicating excellent diagnostic ability. Moderate heterogeneity was detected (prediction area = 0.369; bivariate I(2) = 69.9%). Field of view and retrospective study designs were identified as factors linked with higher specificity. Selection of largest nodes for analysis and participants' age were factors associated with higher sensitivity. The pooled ADC value was 1.272 × 10(-3) mm(2)/s (CI = 1.150-1.394) for benign nodes and 0.874 × 10(-3) mm(2)/s (CI = 0.773-0.974) for metastatic nodes. CONCLUSION: ADC value is a highly accurate non-invasive diagnostic marker for lymph node metastasis in breast cancer. We recommend acquisition of DWI in women less than 51 years old, with the highest b-value of 1000 s/mm(2), repetition time ≥ 8500 ms, field of view ≥ 350 mm, and selection of the largest node for analysis.