ABSTRACT Extracellular vesicles (EVs), including exosomes, circulate in body fluids and carry pathological genomic information. High‐grade serous ovarian carcinoma (HGSOC) is the most common subtype of ovarian cancer, characterized predominantly by copy number variations (CNVs). This study evaluated the clinical significance of EV‐DNA focusing on CNV profiles and exploring its potential as a companion biomarker for predicting therapeutic response and supporting diagnosis in HGSOC. Droplet digital polymerase chain reaction detected concordant CNVs between tumor DNA and EV‐DNA in ascites but not in serum. EV‐DNA showed higher concordance with the CNV status of tumor DNA than cell‐free DNA in ascites. The EV‐DNA amount in malignant ascites was significantly higher than that in ascites from benign ovarian tumors ( p < 0.05), and elevated EV‐DNA was observed even in cytology‐negative ovarian cancer, suggesting an increase early in disease development. Moreover, CNV profiles showed marked differences between responders and nonresponders to poly(ADP‐ribose) polymerase (PARP) inhibitors. An equation based on five genes ( ARID1A, NOTCH3, CSMD3, ELP4 , and BARD1 ) showed strong predictive performance for olaparib response (area under the curve = 0.91). Collectively, these findings indicate that the CNV status of EV‐DNA may serve as a non‐invasive companion biomarker for patient stratification and therapeutic monitoring in HGSOC.
Uekusa et al. (2026) studied this question.