Background
Multicenter observational cohort, N=142 primary ovarian cancer patients on PARP inhibitor maintenance after first-line platinum chemotherapy (CR/PR responders), retrospective monitoring of CA125 during maintenance
Interventions and follow up
Arm A: PARP inhibitor maintenance with CA125 surveillance
Arm B: Standard GCIG biochemical criteria for progression definitio
Primary endpoint: CA125 monitoring utility during PARPi maintenance for early recurrence detectio
mFollow up: Entire maintenance course, retrospective analysi
Arm B: Standard GCIG biochemical criteria for progression definitio
Primary endpoint: CA125 monitoring utility during PARPi maintenance for early recurrence detectio
mFollow up: Entire maintenance course, retrospective analysi
Results
RECIST-defined progression: 38.03% (54/142 patients); GCIG criteria: 20.42% (29/142)
New criterion (CA125 ≥3× nadir): Sensitivity 79.63%, specificity 98.86%, PPV 97.73%, NPV 88.78%, diagnostic accuracy 91.55% (vs GCIG 82.39%)
New criterion (CA125 ≥3× nadir): Sensitivity 79.63%, specificity 98.86%, PPV 97.73%, NPV 88.78%, diagnostic accuracy 91.55% (vs GCIG 82.39%)
Adverse events
Main adverse events: No significant toxicity data (observational study of monitoring strategy). Primary focus on detection methodology.
Conclusions
GCIG biochemical criteria miss 46.3% of progressing patients during PARPi maintenance. New CA125 ≥3× nadir criterion improves sensitivity and NPV while maintaining high specificity, offering practical approach for clinical implementation.
Key Limitations
Key Limitations: Retrospective single-institution design, small sample (N=142). CA125-centric approach; RECIST imaging standard has gaps. Does not address optimal timing of therapy switching or impact on OS.
Clinical Context
Improved progression detection during PARPi maintenance informs treatment switch decisions. GOG-developed criteria highlight need for refined monitoring; supports risk-stratified surveillance strategies in ovarian cancer maintenance.
References
References: Piątek S et al, Oncology Research, 2025