Behavior of Serous Borderline Ovarian Tumors With and Without Micropapillary Patterns: Results of a French Multicenter Study
- First Online:
- Cite this article as:
- Fauvet, R., Demblocque, E., Morice, P. et al. Ann Surg Oncol (2012) 19: 941. doi:10.1245/s10434-011-2039-y
- 216 Views
Behavior of serous borderline ovarian tumors with micropapillary patterns (MP-SBOT) is thought to be worse than those without micropapillary patterns, but few cohort studies have compared epidemiological characteristics, surgical management, and recurrence rates between these two groups.
In a French retrospective multicenter study of 475 borderline ovarian tumors (BOT) treated from 1990 to 2009, we studied patients with a serous BOT and treated after 2000 including 20 patients with and 77 patients without micropapillary patterns.
Patients with MP-SBOT were younger (P = 0.01), often asymptomatic (P = 0.04), and with abnormal CA 125 serum levels (P = 0.04). Peritoneal implants were more frequently observed in these patients (P = 0.01); also, they underwent conservative treatment more frequently (P = 0.002), had a higher risk of misdiagnosis with invasive carcinoma by intraoperative histology (P < 0.05), and had more frequent restaging surgery (P = 0.001). No difference in recurrence was noted between the groups. No disease-related mortality was observed.
Patients with MP-SBOT represent a heterogeneous population in terms of presence of invasive peritoneal implants. Conservative surgery could be a suitable option for MP-SBOT patients without implants and who wish to conserve childbearing potential, without increasing the risk of recurrence.
In the early 1970s, tumors of the ovary were classified as benign, borderline (low malignant potential), or carcinomas according to the degree of epithelial proliferation.1,2 Although this classification has been demonstrated to be correlated with prognosis for benign and malignant tumors, differences in behavior have been observed for borderline ovarian tumors (BOT) suggesting a heterogeneity among this group of neoplasms.3–6
Seidman and Kurman7 showed that serous BOTs exhibiting micropapillary patterns (MP-SBOT) seemed to have a disease course similar to low-grade carcinomas. Subsequently, several studies have underlined the poor prognosis of MP-SBOTs compared with those without this histological feature.8–14 However, bias linked to patient selection might explain the observed differences in behavior. Bristow et al. studied a series of 26 patients with MP-SBOT whose tumor recurred from a larger group of patients with micropapillary patterns, that is, a subgroup of patients with worse prognosis.9 Similarly, Deavers et al. in a series of 99 patients with advanced stages of serous BOT demonstrated a poorer prognosis for the 18 patients with micropapillary patterns.14 However, these authors report a high incidence of both bilateral tumors and invasive implants in the population explaining the unfavorable behavior. All these considerations may lead to aggressive management of MP-SBOT including radical surgery and adjuvant therapies. Therefore, there is a need to clarify the management of MP-SBOT according to the risk of recurrence and development into invasive tumor.
The aims of this multicenter retrospective cohort study were to evaluate whether serous BOT with or without micropapillary patterns had specific epidemiological characteristics, surgical management, and risk of recurrence.
Patients and Methods
From January to December 2010, we conducted a French retrospective multicenter study of 475 women with BOT treated from 1990 to 2009 in 16 institutions (2 cancer centers and 14 specialized gynecology units). To avoid bias linked to histological definition of micropapillary patterns, which was defined by Burks et al. then by Seidman and Kurman in 2000, only 97 patients with a serous BOT and treated after this date were included.5,7 The protocol was approved by the Ethics Committee of the Collège National des Gynécologues et Obstétriciens Français (CNGOF).
The medical records were reviewed to determine age, tumor stage, histology, tumor size, initial surgical procedure, restaging surgery, and the final histological status. Outcome was obtained from the outpatient records.
Histological typing was performed essentially according to the recommendations of the 2003 World Health Organization classification schemata.15 The histological criteria used for the diagnosis of BOT included: (1) stratification of the epithelial lining of the papillae, with formation of microscopic papillary projections or tufts arising from the epithelial lining of the papillae, (2) nuclear atypia, (3) mitotic activity, (4) intracystic clusters of free-floating cells, and (5) absence of stromal invasion. In accordance to previously published criteria, a MP-SBOT was defined as a serous BOT that contains at least 1 area of uninterrupted micropapillary growth measuring more than 5 mm in the maximum dimension and lacking stromal invasion.16 No central pathology review was performed.
Surgical treatment was considered to be conservative if at least a portion of 1 ovary and the uterus were conserved. Conservative ovarian treatment consisted of unilateral cystectomy, unilateral salpingo-oophorectomy (USO), USO and contralateral cystectomy, or bilateral cystectomy. Surgical treatment was considered radical when bilateral salpingo-oophorectomy was performed. In addition to this conservative or radical surgery, careful inspection of all peritoneal surfaces and peritoneal washing, multiple random or oriented biopsies, and infracolonic omentectomy were required to consider the surgery a complete staging operation. Initial surgery was considered to represent an incomplete staging operation in all other cases, regardless of the radical or conservative nature of treatment. Dissection of the pelvic and retroperitoneal lymph nodes was not required for complete staging, as this procedure was abandoned in 1995 as part of the systematic staging of BOTs. Disease staging was performed according to the recommendations of FIGO. Restaging operation was defined as surgical procedures performed after initial incomplete staging, regardless of the initial disease stage, such that (1) the interval between initial surgery and restaging was less than 6 months and (2) the patient had not received adjuvant therapy. The restaging procedure involved laparotomy or laparoscopy, depending on the participating centers.
Statistical analysis was performed using the chi-square test and the t test. P values less than 0.05 were considered an indication of significance.
Epidemiological characteristics, circumstances of diagnosis, and serum tumor markers levels of patients with serous borderline ovarian tumor (BOT) with and without micropapillary patterns
Serous BOT with micropapillary patterns N = 20
Serous BOT without micropapillary patterns N = 77
39.2 ± 16
50.5 ± 17.4
Previous ovarian surgery (%)
Family history of cancer (%)
Symptomatic lesion (%)
Clinical or sonography systematic examination (%)
Tumor size (cm)
8.9 ± 5.3
10.2 ± 5.4
CA 125 > 35 (%)
CA 19.9 > 35 (%)
Comparison of initial surgical management of patients with serous BOT with and without micropapillary patterns
Serous BOT with micropapillary patterns N = 20
Serous BOT without micropapillary patterns N = 77
Type of initial surgery
Type of conservative treatment
Invasive carcinoma (%)
Initial FIGO stages
Complete staging (%)
Invasive implants (%)
Noninvasive implants (%)
No relation was found between abnormal CA125 serum levels and the presence of implants according to micropapillary pattern and tumor size. Of the 16 patients with preoperative evaluation of CA125 serum levels and with MP-SBOT, 13 (81%) had abnormal CA125 serum levels including 6 patients with peritoneal implants. Of the 50 patients without MP-SBOT, 26 (52%) had abnormal CA125 serum levels including 3 patients with peritoneal implants. No difference in complete initial staging was observed between patients with or without abnormal CA125 serum levels independently of micropapillary patterns.
Restaging Surgery and Recurrences
Comparison of restaging surgery in patients with serous BOT with and without micropapillary patterns
Serous BOT with micropapillary patterns N = 20
Serous BOT without micropapillary patterns N = 77
Restaging surgery (%)
Type of final surgery
FIGO final stages
Among the 16 patients with serous BOT without micropapillary patterns who underwent restaging surgery, 12 had a presumed stage IA, 2 a stage IB, 1 patient a stage IIB, and the last a stage IIIC. Of the 16 patients, 4 (25%) were upstaged: 2 of the 12 patients (16%) with stage IA were upstaged to a final stage IIB and IIC, 1 with an initial stage IB to IIC, and the last patient with initial IIB stage to IIIA.
Among the 12 patients with MP-SBOT who underwent restaging surgery, 6 patients had a presumed stage IA, 2 stage IB, 2 stage IC, 1 patient stage IIC, and the last stage IIIA. Of the 12 patients, 3 (25%) MP-SBOT were upstaged: 2 of the 6 patients (33%) with initial stage IA to IC and IIC and the last patient from stage IB to IIC.
Implants were more frequently observed in patients with MP-SBOT (P = 0.01), but no difference was observed between the groups in the rate of invasive implants (Table 2). No additional patients were diagnosed with invasive implants and invasive cancer after restaging operation.
The mean follow-up of the whole population was 15.6 ± 18.5 months. Mean follow-up in patients with serous BOT with and without micropapillary patterns were 29.7 ± 24.4 and 12.2 ± 14.9 months, respectively (P < 0.001). None of the patients received adjuvant therapy.
Recurrence was observed in 1 patient in each group. One patient with MP-SBOT initially treated by cystectomy and contralateral ovarian biopsy (initial stage IB) underwent a restaging surgery with a salpingo-oophorectomy on the ovary initially treated by cystectomy and a contralateral cystectomy and peritoneal biopsies showing noninvasive implants (final IIC FIGO stage). This patient recurred 40 months later on the remaining ovary with the same histology. One patient with serous BOT without micropapillary patterns initially treated by USO and contralateral cystectomy with noninvasive implants recurred 6 months later on the remaining ovary and underwent a second cystectomy. No invasive recurrence was observed in the whole population. There was no disease-related mortality.
The present retrospective multicenter study demonstrates that MP-SBOTs have different epidemiological characteristics to those without this histological feature. Moreover, despite a higher incidence of conservative surgery in patients with MP-SBOT, no difference in recurrence rate was observed with patients with serous BOT without micropapillary patterns.
In the current study, differences in epidemiological characteristics were observed between serous BOT patients with and without micropapillary patterns. Those with MP-SBOT were younger than patients without this histology. These results contrast with a similar study published by Chang et al. showing no difference in the mean age at diagnosis according to micropapillary patterns.17 However, Laurent et al.10 reported a mean age of 25 years in a series of 15 patients with micropapillary patterns undergoing conservative management. Deavers et al.14 suggested that patients with MP-SBOT were more frequently bilateral. These results were confirmed by Laurent et al. reporting that 2/3 of patients with micropapillary patterns had bilateral tumors. However, our results are in agreement with those of Chang et al. who show, in a cohort study, no difference in the rate of bilateral tumors according to micropapillary patterns.17 Abnormal CA125 serum levels were more frequently observed in patients with MP-SBOT independently of the presence of peritoneal implants or tumor size, suggesting that this parameter is associated with the histology. To date, no data are available to evaluate the contribution of serum tumor markers in the preoperative staging of patients with MP-SBOT.
Surgical management of patients with MP-SBOT raises specific issues. Indeed, previous studies have underlined that these patients should be considered patients with low-grade serous carcinoma justifying a more aggressive surgical management compared with classic serous BOT in whom conservative management appears to be an adequate alternative to preserve fertility.5,6,11,14,18–21 In the current study, a higher rate of conservative treatment was noted in patients with MP-SBOT compared with those without this histology. This could be explained by epidemiological characteristics of our population showing that patients with micropapillary patterns were younger (mean age of 38 years), justifying surgery aimed at preserving childbearing potential. Moreover, in a series of conservative treatment for MP-SBOT, although a high recurrence rate was noted, Laurent et al.10 demonstrated that 1/3 of the patients went on to become spontaneously pregnant. These results are in agreement with the only prospective randomized study in a population of patients with early-stage BOTs.22 The team recruited 32 patients with a bilateral serous BOT who wished to preserve fertility and showed a significantly increased chance of pregnancy (hazard ratio [HR] = 3.3, 95% confidence interval [95% CI], 1.4–8.0) but nonsignificant earlier disease recurrence (HR = 1.5, 95% CI, 0.6–3.8) in the women who had ultraconservative surgery (bilateral cystectomy) compared with those who had conservative surgery (cystectomy and contralateral oophorectomy). Moreover, Faluyi et al.23 in a recent meta-analysis underlined that this trial was at low risk of bias. Another issue is the contribution of intraoperative histology to manage patients with MP-SBOT. Although the overall accuracy of intraoperative histology is in keeping with those of previous studies showing that about 2/3 of BOTs are adequately diagnosed, it is important to note that intraoperative histology had a risk of misdiagnosing MP-SBOT as a true invasive tumor and hence the possibility of unnecessary radical surgery.24–27 This is particularly important as we observed that these patients had a higher incidence of abnormal CA125 serum levels reinforcing the histological findings. Indeed, Poncelet et al.28 demonstrated that patients with elevated serum tumor markers were more likely to have radical treatment and complete staging. However, our results contrast with those of this study showing that no difference in complete staging was noted between patients with and without abnormal serum tumor markers.28 This apparent discrepancy could be explained by the relatively high rate of patients in our study (30%) who had complete initial staging, while previous studies report an initial complete staging ranging from 4 to 19% depending on the surgical route, the contribution of intraoperative histology, and the experience of surgeons.25,29
Restaging operation is a matter of debate in patients with BOT because of the low incidence of upstaging and the absence of impact on survival. In our study, restaging surgery was more frequently performed in patients with MP-SBOT involving up to 60% of cases. This is in agreement with French guidelines, suggesting that restaging surgery should be performed in patients with micropapillary patterns.30 Although a higher incidence of upstaging was observed in patients with initial IA FIGO stage exhibiting micropapillary patterns, only 25% of patients with micropapillary features benefited from this strategy, which is similar to patients with serous BOT without micropapillary patterns. These results are in agreement with those of previous studies reporting that between 12.3 and 14.8% of patients are upstaged depending on presumed initial stage with a higher benefit for patients with stage IA.25,31,32 Therefore, it appears that a systematic restaging operation in MP-SBOT patients should be considered, taking into account preoperative serum levels of tumor markers, the type of initial surgery (cystectomy or salpingo-oophorectomy), the presence of bilateral tumors, and the patient’s desire to become pregnant.
The main issue for patients with MP-SBOT is the risk of recurrence and disease-related mortality. In the current study, a low recurrence rate of 5% similar to that observed in patients with serous BOT without micropapillary patterns was noted. This low rate might be explained by our short follow-up. However, the mean follow-up of the patients with micropapillary patterns was higher than that of the whole population and similar to those of previous studies focusing on patients with MP-SBOT.6,8,9,11,14,17 Our results are in agreement with those of Slomovitz et al.,13 showing no negative impact of micropapillary patterns on both recurrence rate and disease-related mortality. In a series with a follow-up of 4 years, Shih et al.33 found that abnormal CA125 serum level, advanced stage, age at diagnosis, and invasive implants were associated with decreased disease-free survival and that micropapillary pattern was associated with a lower disease-free survival of 75.9% (95% CI, 55.6–87.8) compared with 94.3% (95% CI, 88.4–97.3) for patients without micropapillary patterns. However, these authors did not distinguish between patients with micropapillary patterns those with and without invasive implants. Indeed, Chang et al.17 reported a low disease-free survival in patients with MP-SBOT and invasive implants. These results are in keeping with those of Deavers et al.14 showing the negative impact of invasive implants in patients with micropapillary patterns. Finally, it appears that the behavior of MP-SBOT is related more to the presence of invasive implants rather than this particular histology. Hence, studies stratifying recurrence and survival according to both micropapillary patterns and the presence of peritoneal implants (invasive and noninvasive) are required to clarify surgical management.
Some limitations of the current study have to be underlined. First, the retrospective nature of this cohort study and the absence of central pathology review cannot avoid bias even if patients of the first period of the study where there was no clear histological definition of micropapillary patterns have been excluded. Second, the relative short follow-up might have led to an underestimation of the rate of recurrence. However, previous studies have underlined that most MP-SBOTs recur within the first 2 postoperative years.6,8,9,11,14,17 Third, because of the sample size, it is difficult to determine the contribution of serum tumor markers in both initial surgical management and in the follow-up to detect recurrences. Finally, further studies are required to identify patients with MP-SBOT associated with a risk of implants, particularly with an invasive component.
In conclusion, our results support that patients with MP-SBOTs have specific epidemiological characteristics but also represent a heterogeneous population mainly related to the presence of invasive peritoneal implants. Patients with MP-SBOT without implants and who desire to preserve childbearing potential could be candidates for conservative surgery without increasing the risk of recurrence. Further large studies in patients with micropapillary patterns and peritoneal implants are required to determine their impact on conservative management and recurrences.
The authors thank the participating centers: Centre Hospitalier Universitaire Amiens (Prof. Gondry), Centre Hospitalier Universitaire Angers (Prof. Descamps), Centre de Recherche et de Lutte contre le Cancer Paul Papin Angers (Dr. Fondrinier), Centre Hospitalier Général Le Mans (Dr. Pillot), Centre Hospitalier Universitaire Lille (Prof. Querleu), Centre Hospitalier Universitaire Marseille (Prof. Blanc), Centres Hospitaliers Universitaires Paris: Hôpital Bichat (Prof. Madelenat), Hôpital Cochin (Prof. Chapron), Hôpital de la Pitié-Salpétrière (Prof. Lefranc, Prof. Dommergues), Hôpital Saint Antoine (Prof. Milliez), Hôpital Tenon (Prof. Uzan), Centre Hospitalier Universitaire Reims (Prof. Graesslin), Centre Hospitalier Général Roubaix (Prof. Querleu), Centre Hospitalier Universitaire Rouen (Prof. Marpeau), Institut Gustave Roussy Villejuif (Prof. Morice), and Centre Hospitalier Universitaire Strasbourg (Prof. Baldauf).