Abstract
Background/Aim: Borderline ovarian tumours (BOTs) frequently affect women of reproductive age. Fertility-sparing surgery (FSS) is commonly offered, but recurrence risk varies by surgical approach and histological subtype, while malignant transformation remains incompletely defined. This systematic review and meta-analysis aimed to quantify recurrence and malignant transformation risks following different surgical strategies for BOTs.
Materials and Methods: A prospectively registered protocol (PROSPERO: CRD420251172155) was followed. Studies of adult women with histologically confirmed BOTs undergoing FSS – unilateral or bilateral cystectomy, unilateral salpingo-oophorectomy (USO), or USO with contralateral cystectomy – or radical surgical staging (pelvic clearance) were included. Random-effects meta-analyses generated pooled risk estimates with 95% confidence intervals (CIs); heterogeneity was assessed using I2. Methodological quality was evaluated using MINORS.
Results: Forty-one studies including 5,364 women were analyzed, with follow-up ranging from 7 months to 15 years. Recurrence following FSS was highest after cystectomy-based procedures: 0.31 (95%CI=0.15-0.47; I2=97.24%) after unilateral cystectomy and 0.41 (0.26-0.57; I2=80.59%) after bilateral cystectomy. Lower recurrence was observed after USO (0.13; 0.10-0.16; I2=81.55%) and USO with contralateral cystectomy (0.31; 0.17-0.45; I2=91.40%). Radical surgery was associated with the lowest recurrence risk (0.03; 0.02-0.04; I2=26.42%). In serous BOTs, recurrence after cystectomy was 0.29 (0.23-0.56) and after USO 0.15 (0.09-0.24). In mucinous BOTs, corresponding estimates were 0.26 (0.14-0.38) and 0.08 (0.02-0.13). Overall malignant transformation occurred in 1% of women (0.01; 0.01-0.02) but increased substantially following recurrence (0.22).
Conclusion: Cystectomy-based FSS is associated with significantly higher recurrence than oophorectomy-based approaches or radical surgery. Although malignant transformation is rare overall, its marked increase following recurrence highlights the need for histology-informed counselling and long-term, risk-adapted surveillance.
- Borderline ovarian tumor
- fertility-sparing surgery
- cystectomy
- salpingo-oophorectomy
- recurrence
- malignant transformation
- serous
- mucinous
- systematic review
- meta-analysis
- review
Introduction
Borderline ovarian tumors (BOTs) represent a distinct subset of epithelial ovarian neoplasms, accounting for approximately 10-20% of epithelial ovarian tumors (1). Histologically, BOTs are heterogenous, and contrary to invasive epithelial ovarian cancer (OC), they are most frequently diagnosed at an early stage, with nearly 80% confined to the ovaries at presentation (2). BOTs mainly affect women under the age of 40 (3) and given the excellent overall survival, fertility-sparing surgery (FSS) has become a common management strategy for women desiring future fertility. Nevertheless, the risk of recurrence remains a major clinical concern, varying significantly based on adequate surgical staging (4), and occasionally histological elements as well as tumor biology (5). For example, serous BOTs (sBOTs) frequently show hierarchical branching, stratification, and non-invasive peritoneal implants, while a subset demonstrates a micropapillary pattern that is associated with higher recurrence risk and a greater likelihood of extra-ovarian spread (6). In comparison, mucinous BOTs (mBOTs) are often large, unilateral tumors, which carry a higher potential for progression to invasive carcinoma, particularly when areas of intraepithelial carcinoma or microinvasion are present (7). To date, surgical management of BOTs largely follows principles established for epithelial ovarian cancer, with comprehensive surgical staging – performed either as radical surgery or fertility-sparing surgery – considered the gold standard.
Despite favorable survival outcomes, studies assessing recurrence and comparative management strategies in BOTs remain limited and occasionally inconsistent. Prior studies and meta-analyses have evaluated recurrence following FSS and radical surgery in BOTs (8-10), generally demonstrating higher recurrence rates with conservative management which has no clear impact on overall survival (9, 10). However, this data is predominantly retrospective, mainly combines heterogeneous surgical approaches and often fails to stratify outcomes by histological subtype or entirety of staging. Moreover, follow-up strategies are highly variable, contributing to either over- or under-surveillance. Also, the risk of malignant transformation – particularly following recurrence – remains incompletely characterized, with reported estimates varying substantially. As surgical practice and fertility-sparing strategies have evolved over the past decade, an updated synthesis of recurrence and malignant transformation risk is needed to better inform patient counselling and long-term management.
To this end, we conducted a systematic review and meta-analysis to quantify the risk of recurrence following fertility-sparing and radical surgical staging in women with BOTs, and to estimate the risk of malignant transformation in both primary and recurrent disease.
Materials and Methods
This systematic review and meta-analysis were conducted according to a prospectively registered protocol (PROSPERO: CRD420251172155) and are reported in accordance with the PRISMA guidelines (11).
Literature search. PubMed, Embase.com, and Embase via Ovid were searched from inception to November 2025 for English-language human studies. The search integrated terms covering the target population, relevant interventions, and key outcomes including ‘borderline ovarian tumor’, ‘borderline ovarian cancer’, ‘bot’, ‘cystectomy’, ‘salpingo-oophorectomy’, ‘oophorectomy’, ‘adnexectomy’, ‘pelvic clearance’, ‘hysterectomy’, ‘conservative surgery’, ‘fertility-sparing’, ‘recurrence’, ‘relapse’, ‘malignant transformation’, ‘progression’, ‘outcome’. The search strategy was validated by confirming that it successfully retrieved four studies known to meet the eligibility criteria based on preliminary searches. Reference lists of included studies were manually screened in addition to electronic search.
Study selection. Two independent pairs of reviewers (AT/SL and EG/SL) assessed studies in a two-step process. Initially, abstracts were screened to identify potential studies for eligibility. Subsequently, full-text articles were examined to determine final inclusion. Any discrepancies were resolved by the senior author (MS). Studies including women aged 18 years or older with histologically confirmed BOTs of any subtype were eligible (Population). Eligible studies provided data on primary outcomes of interest, which were recurrence rates and/or malignant transformation to invasive ovarian carcinoma (Figure 1). Eligible studies were required to provide extractable clinical annotation data, primarily relating to surgical management (Intervention). Interventions included fertility-sparing surgery (FSS) or radical surgery, comprising unilateral cystectomy (UC), bilateral cystectomy (BC), unilateral salpingo-oophorectomy (USO), unilateral salpingo-oophorectomy with contralateral cystectomy (USO+CC), and pelvic clearance. Pelvic clearance was defined as removal of both ovaries and fallopian tubes, with or without hysterectomy. In all cases, omental/peritoneal biopsies, (+/− infra or supracolic omentectomy) and washings supplemented the main intervention to complete staging (Figure 2). Nested cohorts within randomized controlled trials, cohort studies, and case–control studies were included.
Population Intervention Comparison Outcomes’ (PICO) structure of the systematic review and meta-analysis.
PRISMA flow chart illustrating study identification and selection for the systematic review of borderline ovarian tumor management, recurrence, and malignant transformation.
Studies primarily addressing recurrent BOTs, invasive or non-epithelial ovarian cancers, or paediatric and adolescent populations were excluded. Studies were also excluded if they did not report outcomes relevant to surgical approach, recurrence and/or malignant transformation in BOTs, or if they presented mixed populations without separate data for the histological subtype of BOTs. Reviews, editorials, letters, case reports or series, and conference abstracts without full peer-reviewed publication were excluded. Animal studies and basic science research were also excluded.
Comparators. Recurrence and malignant transformation rates were compared between sBOTs and mBOTs and across surgical approaches.
Quality assessment and data extraction. Two independent pairs of reviewers (AT/EG and AL/SL) assessed the methodological quality of all included studies using the Methodological Index for Non-Randomized Studies (MINORS), with any discrepancies resolved by the senior author (MS) (12). External validity was judged based on the transparency and standardization of definitions for BOT diagnoses and outcome measures, as well as whether the study population was representative, for example, through consecutive or randomized recruitment. Studies that failed to meet at least three criteria related to internal or external validity were classified as having a high risk of bias.
Statistical analysis. We performed random-effects meta-analysis to calculate pooled effect sizes for recurrence of BOTs and for malignant transformation, each with corresponding 95% confidence intervals (CIs). When studies reported more than one effect estimate, we used the most fully adjusted measure. Sensitivity analyses were performed to estimate effect size of primary outcomes for each surgical category (UC, BC, USO, USO+CC, and pelvic clearance) or histology (sBOT, mBOT), with separate pooled effect sizes, CIs generated for each group.
Heterogeneity was measured using the I2 statistic with <50% considered low, 50-75% moderate, and >75% substantial heterogeneity. All analyses were conducted using Stata software (version 15.0, College Station, TX, USA).
Accuracy of data. MS/AT served as the guarantors for the integrity and accuracy of the study data. The corresponding and senior author (MS) held final responsibility for the decision to submit the manuscript for publication (6).
Results
Our search yielded 1,920 citations. After title and abstract screening and full-text assessment, 41 studies were included in the final synthesis, comprising a total of 5,364 women (Figure 2).
The included studies were published between 1996 and 2025 and originated primarily from Europe and Asia, with the greatest number of studies conducted in France, Italy, China, and the United Kingdom. Two studies were conducted in the UK, while no eligible studies originated from the United States (Table I).
Baseline characteristics of included cohort studies.
Total study populations ranged from 18 to 1,390 participants. Reported median ages ranged from 28 to 58.5 years, while mean ages ranged from 28.5 to 52.7 years. Most studies reported a central age in the third to fifth decades of life. Overall recurrence in BOT was 0.22 (95%CI=0.25-0.28; I2=38.32%). Across included studies, follow-up was variably reported and highly heterogeneous, ranging from 7 months to 15 years, with a typical median follow-up approximately 5 years (median of reported medians ≈ 61 months, IQR ≈ 45-74 months)
Recurrence rates following FSS. Across 36 studies including 4,913 women, the pooled recurrence rate following FSS with cystectomy (unilateral or bilateral) for any pathology was 0.32 (95%CI=0.25-0.40, I2=84.50%). Based on 31 studies involving 4,366 women, FSS with unilateral (only) cystectomy demonstrated a pooled recurrence estimate of 0.31 (95%CI=0.15-0.47, I2=97.24%). Based on 19 studies with 4,366 women, FSS with bilateral cystectomy had a pooled recurrence rate of 0.41 (95%CI=0.26-0.57, I2=80.59%). Thirty-four studies including 4,911 women undergoing FSS with unilateral salpingo-oophorectomy had a recurrence estimate of 0.13 (95%CI=0.10-0.16, I2=81.55%), whereas, 20 studies including 2,802 women showed that FSS with unilateral salpingo-oophorectomy combined with contralateral cystectomy had a pooled recurrence estimate of 0.31 (95%CI=0.17-0.45, I2=91.40%) (Table II).
Number of women, effect sizes, confidence intervals, and heterogeneity (I2) for surgical procedures and borderline ovarian tumor recurrence, overall and by histological subtype (any pathology, sBOT, and mBOT).
Recurrence rates following radical staging surgery. Based on 28 studies including 4,398 women, radical surgery in the form of cytoreductive surgery including pelvic clearance was associated with a pooled recurrence estimate of 0.03 (95%CI=0.02-0.04, I2=26.42%) (Table II).
Recurrence rates in sBOT. Sensitivity analysis in the sBOT group, based on six studies comprising 668 women, showed that cystectomy as part of FSS was associated with a pooled recurrence estimate of 0.29 (95%CI=0.23-0.56, I2=75.84%). Based on five studies including 543 women, USO was associated with a pooled recurrence estimate of 0.15 (95%CI=0.09-0.24, I2=33.33%) and pelvic clearance recurrence (radical surgery) estimate was 0.05 (95%CI=0.02-0.09, I2=22.09%) (Table II).
Recurrence rates in mBOT. In the mBOT group, based on four studies including 641 women, recurrence estimates following FSS with unilateral cystectomy were 0.26 (95%CI=0.14-0.38, I2=0%) and following USO 0.08 (95%CI=0.02-0.13, I2=62.39%). Based on two studies (75 female patients), recurrence following radical staging including pelvic clearance was 0.03 (95%CI=0.008-0.12, I2=0%) (Table II).
Malignant transformation of borderline ovarian tumours. Based on 13 studies including 2,050 women, malignant transformation rate was 0.01 (95%CI=0.01-0.02, I2=66.46%). In the case of recurrent BOTs this was 0.22 (95%CI=0.25-0.28, I2=38.32%).
Quality assessment. Across 40 included studies which were quality assessed, methodological quality by MINORS was moderate overall (median total score 10, IQR=8-16). Comparative studies scored higher (median 17, IQR=16.8-17) than non-comparative studies (median 9, IQR 8-10). Items most consistently adequate were a clearly stated aim and appropriate endpoints (~98%). Major weaknesses were prospective data collection (90% not reported), study size calculation (98% not reported) and reporting of loss to follow-up (55% not reported; 42% inadequate). Figure 3 summarizes MINORS breakdown for each item.
MINORS quality assessment breakdown.
Discussion
Summary of findings. In this systematic review and meta-analysis of 41 studies including over 5,000 women with BOT, we found that although recurrence after initial surgery was considerably common following fertility-sparing procedures, malignant transformation remained rare overall. Approximately one third of women who underwent ovarian cystectomy experienced a recurrence; this affected nearly one in three women after unilateral cystectomy and over two in five after bilateral cystectomy. In contrast, recurrence occurred in approximately one in eight women following USO and in approximately one to-two in five women when USO was performed with contralateral cystectomy.
The lowest recurrence risk was observed after radical surgery, with fewer than one in 30 women experiencing recurrence following pelvic clearance. Recurrence patterns differed by histological subtype. Just over 1-2 in 10 women with sBOT will recur after USO compared with approximately one in 20 after pelvic clearance. Over one in four women with sBOT will recur after cystectomy. In contrast, women with mBOTs had lower recurrence rates overall, with recurrence occurring in approximately one quarter after cystectomy and fewer than one in 10 following USO.
Malignant transformation was uncommon, occurring in approximately one in 100 women overall. However, among women who experienced a recurrence, more than one in five subsequently developed invasive disease, highlighting the clinical importance of recurrence prevention and long-term surveillance.
Quality of the included studies remained variable, and highly heterogeneous. This can be an indication of potentially significant variation in clinical management of BOTs across different centers.
Interpretation. Our findings are keeping with the existing published evidence that FSS is oncologically safe, but is associated with a higher risk of recurrence compared to radical surgery (2, 53, 54). Importantly, the magnitude of this risk is strongly influenced by the extent of ovarian preservation and perhaps by histological subtype. Across all fertility-sparing approaches, cystectomy – particularly bilateral cystectomy – was associated with the highest recurrence rates, affecting approximately one third of women overall and over two fifths following bilateral procedures. In contrast, USO was associated with a markedly lower recurrence risk, approaching that observed after more radical procedures. These findings are consistent with previous observational studies and meta-analyses (2, 55), but the present analysis offers greater granularity by separating surgical strategies rather than grouping fertility-sparing approaches as a single entity. This distinction has important implications for counselling, as the oncological risk associated with cystectomy appears substantially greater than that associated with oophorectomy-based fertility preservation.
Histological subtype emerged as a further modifier of recurrence risk. Women with sBOTs demonstrated higher recurrence rates across FSS compared with those with mBOTs. This observation aligns with known biological and pathological differences between subtypes. Serous tumors are more frequently bilateral, may be associated with non-invasive peritoneal implants, and can exhibit micropapillary architecture, all of which have been associated with increased recurrence. In contrast, mBOT are typically unilateral and anatomically confined, which may partly explain the lower recurrence rates observed after USO in this subgroup. Nevertheless, lower recurrence rates in mBOTs should not be interpreted as uniformly lower oncological risk. Mucinous tumors are more likely to harbor foci of intraepithelial carcinoma or microinvasion, and progression to invasive carcinoma may occur without multiple borderline recurrences (5). Our findings therefore support histology-specific counselling, recognizing that recurrence frequency and malignant potential may not follow identical trajectories across subtypes.
A clinically important observation is the divergence between the overall risk of malignant transformation and the markedly higher risk observed following recurrence; this had been previously reported but the evidence was slim (57). While malignant transformation was uncommon overall, occurring in approximately 1% of women, more than one in five women who experienced recurrence subsequently developed invasive disease. This suggests that recurrence represents a clinically meaningful inflection point reflecting a rather more persistent biological behavior, which could harbor risk of malignancy. Our data support a more cautious approach to recurrent disease, particularly in women who have completed childbearing, with consideration of definitive surgery including bilateral oophorectomy and hysterectomy (57), which has been supported by a recent Italian study. This should also be taken into account when planning the scope of follow up in the recurrent disease setting, which must take into account the use of serum biomarkers (Cancer antigen 125, Cancer antigen 19-9, Carcinoembryonic antigen) as a sensitive indicator of malignancy (27).
Finally, the substantial heterogeneity observed across studies highlights potential variation in surgical practice, staging completeness, pathological reporting, and follow-up strategies. This variability likely contributes to the wide range of recurrence estimates reported in the literature and underscores the need for more standardized management pathways. For example, there is still lack of consensus as to whether omental biopsy is sufficient or a infra/supra colic omentectomy is needed for staging purposes.
Integration of molecular biomarkers with surgical risk stratification. Although molecular data were not directly assessed within this meta-analysis, emerging evidence suggests that molecular profiling may help contextualize recurrence risk following FSS. sBOTs frequently harbor alterations in the MAPK pathway, most commonly involving BRAF or KRAS mutations. BRAF-mutated tumors have been associated with indolent behavior and lower risk of progression, whereas KRAS mutations may be associated with higher recurrence, risk of non-invasive implants and further progression to low-grade serous carcinoma (5). Serum biomarkers such as Cancer Antigen 125 retain limited diagnostic value but may assist in post-operative surveillance, particularly in serous disease when elevated pre-operatively (secretors). At present, molecular biomarkers should be regarded as adjunctive tools rather than determinants of surgical decision-making, but their integration into future prospective studies may allow more refined, biology-driven counselling in women considering FSS.
Recommendations. FSS should remain an accepted management option for appropriately selected women with BOTs, particularly those with early-stage disease that wish to preserve fertility. However, counselling must be explicit regarding the differential recurrence risks associated with specific surgical techniques.
Where fertility preservation is desired, USO should be the standard approach over cystectomy whenever technically and oncologically feasible, given its substantially lower recurrence risk. Cystectomy – especially bilateral – should be reserved for carefully selected cases and undertaken with clear discussion regarding the likelihood of spillage and future recurrence and the need for prolonged surveillance.
Histological subtype should be incorporated into counselling and follow-up planning. Women with sBOTs should be counselled regarding their higher propensity for recurrence, while women with mBOTs should be informed that lower recurrence rates do not eliminate the risk of malignant foci.
In the recurrent setting, particularly following FSS, clinicians should adopt a lower threshold for radical cytoreductive staging including hysterectomy. Once family planning is complete, completion surgery – hysterectomy with contralateral or bilateral salpingo-oophorectomy – should be actively discussed. Nevertheless, hysterectomy itself has been a topic of debate and should be carefully considered in selected cases. Discussions around this, should be individualized (26) and taking into account the timeframe of potential recurrence (58). Of the recurrences, 37% are diagnosed within the first 2 years, 31% within year 2-5, 32% after year 5, 10% more than 10 years (1). Our data spanned across a follow up period ranging significantly from 7 months to 15 years. Further to this, there was no option of time to event analysis to further comment on recurrence timing was a limiting factor in our further interpretation.
Conclusion
Prospective, multicenter studies with standardized definitions of surgical approach, staging adequacy, recurrence, with long follow up to allow assessment of malignant transformation are needed to reduce heterogeneity and improve risk estimation. Future research should distinguish between ovarian-confined recurrence and extra-ovarian or invasive relapse, as these likely represent biologically distinct entities.
Integration of molecular profiling into prospective cohorts represents a key opportunity to refine risk stratification. Combining surgical variables, histology, and molecular features may enable safer de-escalation of treatment in biologically low-risk disease while identifying patients who may benefit from earlier definitive surgery (59).
These findings support a personalized management strategy integrating surgical approach, tumor biology, and patient priorities, and provide a robust evidence base to inform counselling, surveillance, and future research. Follow-up strategies should be individualized according to surgical approach, histological subtype, and recurrence history, rather than applying uniform surveillance protocols to all women with BOTs.
Footnotes
Authors’ Contributions
Angeliki Taniskidi*: Data curation; Investigation; Methodology; Formal analysis; Writing – original draft; Writing – review & editing; Evangelia Giannas*: Data curation; Investigation; Methodology; Formal analysis; Writing – original draft; Writing – review & editing; Sofia Lekka: Data curation; Investigation; Methodology; Formal analysis; Writing – original draft; Writing – review & editing; Aikaterini Taniskidi: Data curation; Writing – review & editing; Arabella Lecoyte: Quality Assessment, editing of manuscript; Mark Brincat: Editing of manuscript and supervision; James Dilley: Editing of manuscript and supervision; Arjun Jeyarajah: Editing of manuscript and supervision; Alexandra Lawrence: Editing of manuscript and supervision; Elly Brockbank: Editing of manuscript and supervision; Saurabh Phadnis: Editing of manuscript and supervision; Ranjit Manchanda: Editing of manuscript and supervision; Michail Sideris: Conceptualization; Supervision; Project administration; Validation; Writing – review & editing; Guarantors: Angeliki Taniskidi and Michail Sideris.
Conflicts of Interest
The Authors declare no conflicts of interest in relation to this study.
Funding
This study received no specific funding.
Artificial Intelligence (AI) Disclosure
The Authors declare that no generative artificial intelligence was used to write the manuscript.
- Received January 18, 2026.
- Revision received January 31, 2026.
- Accepted February 4, 2026.
- Copyright © 2026 The Author(s). Published by the International Institute of Anticancer Research.
This is an open access article under the terms of the Creative Commons Attribution License, which permits use, distribution and reproduction in any medium, provided the original work is properly cited.
References
In this issue
Jump to section
Related Articles
Cited By...
- No citing articles found.