To assess whether changes in TP53 mutations and copy number alterations (CNA) in plasma circulating tumor DNA (ctDNA) can predict treatment response and prognosis in platinum‐resistant recurrent ovarian cancer (PROC) patients.

Fifty‐seven PROC patients were recruited. Forty‐three patients with matched tumor and plasma samples were analyzed via both a tumor‐informed ctDNA assay (TICA) and a tumor‐uninformed ctDNA assay (TUCA) profiling TP53 mutations and CNA. The TUCA algorithm was optimized based on TICA results. Fourteen patients without matched tumor tissues were used just for TUCA analysis.

A ctDNA decrease of ≥ 80% from baseline or ctDNA negativity during treatment detected by the TICA (defined as favorable TICA changes) strategy before the third cycle predicted the best overall response, with 81.8% sensitivity and 84.6% specificity. The TUCA strategy was defined as a combination of TP53 mutations and CNA changes. A favorable TUCA change before the third cycle predicted the best overall response, with 90.0% sensitivity and 63.2% specificity. In 12 patients without clinical benefit, the median lead time to detect drug resistance from TUCA to the Response Evaluation Criteria in Solid Tumors was 86.0 days. Patients with favorable ctDNA changes (n = 15) detected by TUCA before the third cycle had a median progression‐free survival of 9.2 months, versus 3.6 months in those without (n = 34) (HR: 2.88; 95% CI 1.56–5.30; log‐rank p = 0.0008).

Similar to TICA, ctDNA changes detected by TUCA combined with TP53 mutations and CNA could predict treatment response and prognosis in PROC patients without requiring tumor tissues.