Risk-Adapted Surveillance in Borderline Ovarian Tumours (BOTs): The "Barts" Evidence-Based Framework

Background/Objectives: Borderline ovarian tumours (BOTs) are distinct epithelial neoplasms with excellent survival but variable recurrence, including late relapse and occasional malignant transformation. Follow-up strategies remain inconsistent internationally, with no standardised, risk-adapted framework. We aimed to synthesise the evidence on BOT recurrence patterns, risk factors and surveillance strategies, and to propose a structured, risk-adapted surveillance framework. Methods: This is a narrative expert synthesis rather than a systematic review. Three evidence sources were combined: our previously published comprehensive review of BOTs, our recent meta-analysis of recurrence and malignant transformation, and an appraisal of contemporary international guidelines. Clinicopathological and surgical determinants of recurrence were then mapped onto the available follow-up tools through a three-step approach, and the resulting risk strata, surveillance intervals and total follow-up duration were agreed on by consensus within a single tertiary gynaecological oncology centre (the "Barts Framework"). Results: Recurrence occurs in 3-10% of patients, with up to one-third arising beyond five years. Key predictors include fertility-sparing surgery (particularly cystectomy), incomplete staging, advanced stage, residual disease, and adverse histological features such as micropapillary/cribriform architecture and invasive implants. Three targets for surveillance were identified: early detection of recurrence, detection of malignant transformation, and optimisation of fertility. Transvaginal ultrasound emerged as the cornerstone modality, with cross-sectional imaging and tumour markers applied selectively. Four risk strata were derived, each specifying follow-up intensity, modality, total duration and care setting. Low-risk patients can be managed in decentralised settings (gynaecological units), whereas high-risk groups warrant specialist oversight (gynaecological oncology centres). Patient-initiated follow-up is incorporated to minimise unnecessary interventions. Conclusions: This risk-adapted approach provides a scalable framework to standardise BOT surveillance whilst reducing overuse and maintaining oncological safety. The framework is consensus-based and single-institution in origin, and prospective external validation with long-term outcome data is required before wider adoption. Integration of molecular stratification and artificial-intelligence-assisted imaging represents a key future direction.