Surgery of borderline tumors of the ovary: retrospective comparison of short-term outcome after laparoscopy or laparotomy

Transcription

1 Acta Obstetricia et Gynecologica. 2007; 86: ORIGINAL ARTICLE Surgery of borderline tumors of the ovary: retrospective comparison of short-term outcome after laparoscopy or laparotomy ELIN ØDEGAARD 1, ANNE CATHRINE STAFF 1,3, ANTON LANGEBREKKE 1, VIBEKE ENGH 2 & MATHIAS ONSRUD 1,3 1 Department of Obstetrics and Gynecology, 2 Department of Pathology, Ullevål University Hospital, and 3 Faculty of Medicine, University of Oslo, Oslo, Norway Abstract Background. Laparoscopic management of borderline ovarian tumors is controversial. Objective. To retrospectively compare outcome after surgery by laparoscopy or laparotomy for borderline tumors. Methods. Ovarian tumors from all women operated at Ullevål University Hospital during a five-year period were re-evaluated histologically. Patients with borderline FIGO (International Federation of Gynaecology and Obstetrics) stage I tumors were retrospectively compared regarding surgery outcome following laparoscopy or laparotomy. Results. Histological re-evaluation revealed only 3 misclassifications in 608 patients. Borderline tumors represented 36% of epithelial ovarian malignancies. The 107 borderline stage I included 52 serous, 53 mucinous, and 2 endometrioid tumors. Thirty-eight patients were operated on primarily by laparoscopy and 69 by laparotomy (including 14 women starting with laparoscopy). In the laparoscopy group, more women were premenopausal (63% versus 35%, p0.01) and median tumor diameter was smaller (8.6 versus 16.4 cm, p B0.001) as compared to the laparotomy group. When tumor diameter exceeded 10 cm, intraoperative tumor rupture was significantly more frequent during laparoscopy than during laparotomy (p0.01). Less postoperative complications were seen after laparoscopic operations (p0.034), but laparoscopic surgeries were less extensive, without hysterectomy, as compared to laparotomy. During the 1478 months follow-up time, no relapse occurred in either group. After fertility-sparing surgery, there was no statistical significant difference regarding successful pregnancies between the two groups. Conclusions. Laparoscopic treatment of borderline ovarian tumors is feasible if tumor is of moderate size (diameter below 10 cm), gives fewer complications, and shorter hospital stay. Long-term follow-up of larger materials is needed to determine the ultimate recurrence risk as well as fertility rates. Key words: Ovarian neoplasms, laparoscopy, laparotomy, complications Abbreviations: BOT: borderline tumor of the ovary, RMI: risk of malignancy index, FIGO: International Federation of Gynaecology and Obstetrics (Federation Internationale de Gynécologie et d Obstétrique) Borderline ovarian tumors, also called ovarian tumors with low malignant potential, comprise 15% of all epithelial ovarian neoplasms (1). The outcome for women with borderline tumors is much better than for women with invasive ovarian cancer, the 10-year survival rate reported between 77% and 99% (2,3). For FIGO stage I borderline, the 10-year survival rate is excellent, 90 99% (2,3). Borderline tumors affect a younger age group than do invasive epithelial carcinomas, and future fertility is an important issue for many of these women. The same FIGO staging system is used for borderline as for invasive ovarian cancer (3,4). For epithelial ovarian carcinomas, FIGO recommends a comprehensive surgical staging to be done through a generous vertical midline laparotomy incision. The operation should include a total hysterectomy, bilateral salpingo-oophorectomy, abdominal washing Address for correspondence: Elin Ødegaard, Department of Obstetrics and Gynecology, Ullevål University Hospital, Kirkeveien 166, 0450 Oslo, Norway. elin.odegaard@uus.no (Received 27 November 2006; accepted 9 February 2007) ISSN print/issn online # 2007 Taylor & Francis DOI: /

2 Surgery of borderline tumors 621 with cytology, biopsies of peritoneal surfaces, omentectomy, and retroperitoneal lymph node sampling from the pelvic and para-aortic regions. For borderline tumors, the role of lymph node sampling is controversial due to the good prognosis of this disease and the potential morbidity associated with lymph node sampling. In France, lymph node sampling is no longer recommended in borderline ovarian tumors (5). A multicenter study from the USA also concluded that routine pelvic and paraaortic lymph node dissection is unnecessary for the majority of women with ovarian borderline tumors (6). For fertile women who have a pregnancy desire, conservative surgery has been demonstrated as feasible, provided that careful follow-up is given (7). Conservative surgery is defined as preservation of the uterus and at least a part of one ovary. In such cases, the laparoscopic approach could be preferable. The purpose of this study was to compare retrospectively the short-term outcome after primary operation for stage I borderline tumors, performed by laparoscopy or by laparotomy at a single institution. FIGO stage I ovarian tumor is defined as tumor limited to the ovaries with or without positive peritoneal washing, ascites, or rupture of the capsule (4). In addition, we wanted to reevaluate histologically all ovarian tumors operated on in the same time period, to get institutional data on the occurrence of borderline tumores relative to ovarian carcinomas and benign tumors. Materials and methods All histological specimens from ovarian tumors registered at the Department of Pathology, Ullevål University Hospital, in a five-year period (January 2000 through December 2004) were identified and reviewed by a senior pathologist (VE). Electronically registered diagnosis lists from the Department of Pathology were cross-checked against operation registrations and diagnosis lists from the Department of Gynecology. Histological classification was performed according to the World Health Organization (WHO) guidelines (4). Staging was performed according to the FIGO 1988 classification rules (3). The medical records of all patients who after histological revision were confirmed to have borderline stage I tumors were studied retrospectively. The risk of malignancy index (RMI) was calculated retrospectively by the first author (EØ), according to guidelines (8), see criteria in Table I. The decision to perform surgery by laparotomy or laparoscopy was made by the operating gynecologists, and was based on several not well-defined criteria, including serum concentration of CA125, ultrasound findings, and the surgeons experience with advanced laparoscopy. Women who initially underwent laparoscopy, but during the operation had a conversion to laparotomy, were analyzed in the laparotomy group. Patients who had a cyst drainage performed peroperatively, either intentionally (on a cyst presumed to be benign) or unintentionally, are classified in the tumor rupture group due to potential tumor cell spilling. If the cyst was intentionally punctured inside an endobag in order to reduce tumor size and facilitate removal through the abdominal wall, with no spilling into the abdominal cavity, the patient is classified in the non-rupture group. Follow-up information was obtained from the medical records, supplemented by information from the patient s gynecologist or from the patients themselves if the last consultation was outside the hospital. Follow-up intervals were three months the first postoperative year, six months the second year, and once a year thereafter. Table I. Demographic data for the study groups prior to first operation. Values are shown as median and range (minimum and maximum values) for the continuous variables, or as numbers and percentages Preoperative variables Laparoscopy (n38) Laparotomy (n69) p-value Age (years) 44 (22 78) 53 (18 91) 0.07* Premenopausal 24/38 (63%) 24/69 (35%) 0.01** BMI (kg/m 2 ) 24.8 ( ) 25.4 ( ) 1.0* CA125 (U/ml) 21 (492) 105 (41,634) B0.001* CA U/ml 5/38 (13%) 31/69 (45%) 0.001** RMI200 1/38 (3%) 23/69 (33%) B0.001** *MannWhitney U -test, **Chi-square test. Postmenopausal, amenorrhea1 year, or age50 if hysterectomized. BMI, body mass index; RMI, risk of malignancy index, calculated as UMCA125 (ref. 7); U, ultrasound score (13) depending on multilocularity, solid tumor areas, bilaterality, ascites, and extraovarian tumors; M, menopausal score (premenopausal1, postmenopausal3); CA125, absolute serum value (U/ml).

3 622 E. Ødegaard et al. Study approval was given by the Regional Committee for Medical Research Ethics in Eastern Norway. All patients signed an informed consent, which included permission for the authors to contact other hospitals, their private doctor, and also to be contacted personally by phone for follow-up data. Statistical analysis Statistical analysis was performed using Statistical Package for the Social Sciences (version 14.0; SPSS Inc, Chicago, IL). As most variables were not normally distributed, numerical results are presented as medians (and ranges). Numerical variables were analyzed by the Mann Whitney exact U-test (MW). Categorical variables were analyzed by Chi-square test or Fisher s exact test (when the expected values were less than five in a 22 table). A probability level of B0.05 was considered statistically significant. Results In the five-year study period, 608 patients with ovarian tumors were operated on at our institution. Two women with pseudomyxoma peritoneii originating in the appendix were not included in the study group, neither were three patients with another gynecological cancer removed at the same time as the ovarian tumor. During re-evaluation, only 3 of the 603 histological diagnoses were reclassified (Figure 1). The reclassifications included one tumor with primary diagnosis of carcinoma that was reclassified as advanced stage borderline, one tumor reclassified from benign serous cystadenoma to serous stage I borderline, and one tumor reclassified from stage I borderline tumor to dilated fallopian tube. Of the 603 ovarian tumors operated on and reevaluated, 19% are classified as borderline tumors and 33% as carcinomas, the remaining 48% as benign ovarian tumors. Among the malignant ovarian tumors, 36% were borderline tumors. Ovarian tumors operated evaluation 2. evaluation Benign BOT Stage I Stage >I Carcinoma 200 BOT Benign Stage I Stage >I Carcinoma serous 53 mucinous 2 endometroid Intended treatment Laparoscopy Conversion to laparotomy 14 Laparotomy 1. operation Laparoscopy 38 Laparotomy operation (staging) Laparoscopy Laparotomy Laparoscopy Laparotomy Figure 1. Flow chart illustrating retrospectively the course of 603 women operated on for ovarian tumor during at Ullevål University Hospital. 1. evaluation represents the standard postoperative histological diagnosis, whereas the 2. evaluation represents diagnosis after re-evaluation of all tumor specimens by a senior pathologist (VE). The lower part of the flow chart illustrates which operation method was planned ( Intended treatment ) and performed (1. operation) in the 107 women with FIGO stage I borderline ovarian tumor. In some patients, a second (staging) operation was performed ( 2. operation ). Stage, FIGO stage; BOT, borderline ovarian tumor.

4 Surgery of borderline tumors 623 The study group, comparing outcome after laparoscopy and laparotomy, consisted of 107 women with stage I borderline tumors, representing 95% of all cases with borderline tumors in our material. There were 52 serous borderline ovarian tumors (two of these presenting focal invasions less than 3 mm), 53 mucinous borderline ovarian tumors (two of which had focal invasion less than 3 mm), and finally two endometrioid borderline ovarian tumors. In our study group of 107 women, 38 (36%) were primarily operated by laparoscopy only and 69 (64%) were operated by laparotomy. In the latter group, 14 patients were primarily operated by laparoscopy, but had a peroperative conversion from laparoscopy to laparotomy, most often because cancer was suspected. This gives a laparotomy conversion rate of 27% (14 of 52) during the primary operation. As shown in Figure 1, a second operation for staging purposes was performed in 21 of the patients, 17 from the laparoscopy group (13 reoperated on by laparoscopy and 4 by laparotomy), and 4 from the laparotomy group (2 operated on by laparoscopy and 2 by laparotomy). Thus, in total, 73 patients were operated by laparotomy (69 plus 4) and 34 patients by only laparoscopy (38 minus 4). For the 73 patients undergoing laparotomy (first or second operation, Figure 1), the surgery was more extensive than for the 34 patients only operated by laparoscopy. No patients in the laparoscopy group underwent hysterectomy, as compared to 52 of 73 in the final laparotomy group. Twenty-one patients (of 34) in the laparoscopy group underwent bilateral salpingo-oophorectomy, as compared to 60 (of 73) patients in the laparotomy group. Preoperative variables Preoperative variables are summarized in Table I. In the whole study group, 41% of the women were premenopausal, 4% perimenopausal, and 55% postmenopausal. More women were premenopausal in the laparoscopy group as compared to the laparotomy group. Body mass index was similar for the two groups. Median serum CA125 level was significantly higher in the laparotomy group compared to the laparoscopy group, as was RMI, retrospectively scored. RMI above 200 was found in 23 out of the 69 patients who underwent laparotomy in the first operation (Figure 1), and in only one patient of the 38 patients operated by laparoscopy. This patient was 64 years old, the intraoperative frozen section histology was interpreted as benign by the pathologist, but final histology of paraffin-embedded sections showed a serous borderline tumor. Perioperative variables Findings during primary operation are presented in Table II. The distribution according to substages of stage I borderline tumors (IaIc) was not significantly different between the two treatment groups. Median tumor size was 8.6 cm in the laparoscopy group and 16.4 cm in the laparotomy group (p B0.001). In our study group of 107 patients, we had tumor rupture, with or without peroperative intra-abdominal spilling, in 29% of the laparoscopy group, and in 16% of the laparotomy group. This difference was, however, not statistically significant (p 0.2). For stage I borderline tumors exceeding 10 cm in diameter, tumor rupture was significantly more frequent during laparoscopy as compared to during laparotomy (5/8 versus 9/44, p0.014). Three patients in the laparotomy group, who presented with tumor rupture before operation, were not included in the analysis. Postoperative variables There were more postoperative complications (18%) in the laparotomy group as compared to Table II. Perioperative data for the study groups for the primary operation. Values are shown as median (and range, minimum and maximum values, for tumor diameter size), or as numbers and percentages Peroperative findings Laparoscopy (n38) Laparotomy (n69) p-value Substage Stage Ia ** Stage Ib 3 3 Stage Ic Tumor diameter size (cm) 8.6 (430) 16.4 (335) B0.001* Tumor rupture 11/38 (29%) 11/69 (16%) 0.2** Tumor rupture, if tumor diameter10 cm 5/8 (63%) 9/44 (21%) 0.014** Ascites 1 (3%) 12 (17%) 0.03*** Blood transfusion needed *** *MannWhitney U -test, **Chi-square test, ***Fisher s exact test. Substage: Ia, tumor limited to one ovary, capsule intact; Ib, tumor limited to both ovaries; Ic, tumor on ovarian surface, or capsule rupture, or ascites/peritoneal washing containing malignant cells.

5 624 E. Ødegaard et al. the laparoscopy group (3%), and wound hematoma was the most frequent complication (Table III). However, laparoscopic surgeries were less extensive, without hysterectomy, as compared to laparotomy. Median hospital stay was also longer for patients operated on by laparotomy: 5 days versus 1.7 day. Following the primary operation, 21 patients (17 in the laparoscopy and 4 in the laparotomy group) underwent a second operation for staging, as they were not completely clinically staged at first operation (see Figure 1). After the staging operation for these 21 patients, only one patient was upstaged from 1a to 1 b, because the contralateral ovary was also affected. She was accordingly operated on, the remaining adnex, uterus, and omentum being removed. Follow-up time was from 14 to 78 months, with no relapse occurring in either group. One 82-yearold woman in the laparotomy group was followed for seven months, where after she died of an intercurrent disease. In the laparoscopy group, there was a higher baby take-home rate as compared to the laparotomy group: 6 versus 4 live-birth deliveries. This difference was not statistically significant. At first operation 31 women were younger than 40 years. Of these 26 were operated on by fertility-sparing surgery, leaving one adnexa and uterus intact, 13 in the laparoscopy group and 13 in the laparotomy group. Diagnostic accuracy of peroperative frozen section histology Thirty of the 107 borderline tumors were evaluated peroperatively by frozen section histology. When compared with the final histology, 18 of these 30 tumors (60%) were erroneously diagnosed as benign. Ten out of 14 mucinous borderline tumors were misdiagnosed as benign, whereas among 16 serous borderline tumors, two frozen sections were considered benign and two were considered as inconclusive. Discussion The Gynecological Department at Ullevål University Hospital is a primary referral institution for the city of Oslo, covering approximately 550,000 inhabitants. Also, patients form neighboring regions are referred if advanced ovarian cancer is suspected. Except for a few benign cysts operated on at private clinics, and some ovarian carcinomas operated on at a neighboring cancer center (The Norwegian Radium Hospital), nearly all primary ovarian tumor surgery for Oslo patients is performed in our institution. In other populations, borderline tumors of the ovary have been reported to represent 420% of all epithelial malignancies (9,10). The higher fraction of borderline tumors found in our material (36%) is possibly due to a fraction of our patient population with advanced ovarian cancer being operated on elsewhere. In our study, only 5% of the borderline tumors were of stage more than I (Figure 1, 2. evaluation), whereas other studies report a higher proportion of more advanced borderline stages, % (2,5). This discrepancy may also be due to leakage of more advanced cases to The Norwegian Radium Hospital. In view of the excellent prognosis and many patients of young age, the type of surgery becomes an important issue. Adnexal masses appearing benign preoperatively and peroperatively are nowadays treated by laparoscopy in most industrialized countries. Perioperative examinations do not always distinguish among benign, borderline, and earlystage carcinomas. As borderline ovarian tumors more often arise in young women, in whom malignancy is less common, their initial surgery is often Table III. Total postoperative data for the study groups following first operation and a potential second staging operation (performed in 21 patients, see Figure 1). Values are shown as median (and range, minimum and maximum values, for days of hospital stay), or as numbers and percentages Postoperative findings Laparoscopy only a (n34) Laparotomy b (n73) p-value Total complications 1/34 13/ * Hematoma in the top of vagina 0 3 Wound hematoma 1 6 Bladder or bowel injury 0 2 Fascia rupture 0 2 Median total hospital stay (adding first and a possible second staging operation) (days) 1.7 (0.54) 5 (136) 0.001** *Fisher s exact test, **MannWhitney U-test. a The laparoscopy group (n34): 38 patients operated on primarily by laparoscopy minus 4 patients with a staging laparotomy as second operation. b The laparotomy group (n73): 69 patients operated on primarily by laparotomy plus 4 patients (primarily operated by laparoscopy) with a staging laparotomy as second operation.

6 Surgery of borderline tumors 625 laparoscopy (11). Laparoscopy offers the benefits of better lighting of the abdominal cavity with better views of the peritoneal surface and the diaphragm. Some of the disadvantages include the lack of tactile sensation, more difficult manipulation, and removal of larger masses through the abdominal wall, and thereby a higher risk of tumor rupture and intraabdominal spilling (12). This is of special concern for mucinous tumors, with a risk of pseudomyxoma peritoneii formation. In our study, there was a significant difference in intraoperative tumor rupture in the laparoscopy group compared to the laparotomy group when tumor diameter exceeded 10 cm, similarly to other studies (5,13). Maneo et al. recommend that laparoscopic treatment of borderline ovarian tumors should be reserved for tumor diameter of 5 cm or below (13). Intraoperative tumor rupture will upstage malignant cysts from stage Ia to stage Ic (4). Vergote et al. confirmed that tumor rupture prior to surgery is an independent poor prognostic variable in ovarian carcinomas (14). It is suggested that tumor rupture before surgery may be related to the tumor biology, as aggressive growth and tumor fragility, with augmented risk of tumor rupture, are related factors. On the other hand, several other studies have shown that cyst rupture during surgery in invasive ovarian cancer is not associated with reduced survival (1517). Despite the fear of many gynecologists that tumor rupture during surgery may promote metastasis and thereby hasten patient death, there is no convincing evidence that this fear is justified (14,18). Intra-abdominal tumor spilling of borderline tumor cells may be a risk factor for port-site metastasis. In order to avoid spilling and to protect the abdominal wall, the use of an endobag might be of clinical importance. In this study, we did not observe any port-site metastasis or other tumor recurrences. Longer follow-up of larger patient cohorts is necessary to evaluate this risk further. Today, several surgical centers advocate laparoscopy for staging and treatment of both borderline tumors and early stages of ovarian cancer (19,20), but it is still generally agreed that patients with invasive ovarian cancer should undergo laparotomy to ensure optimal staging (21). Complete staging procedures, including salpingo-oophorectomy, omentectomy, peritoneal washing, and biopsy taking can be performed laparoscopically. The risk of encountering an unexpected ovarian cancer in the course of laparoscopy for an adnexal mass will depend on the population treated. Important selection criteria are menopausal status, ultrasound findings, and serum CA125 level. Neither RMI score derived from these parameters nor CA125 determination alone can be used as a discriminator between ovarian carcinoma and borderline tumor (all stages). This is also illustrated in our study, where only 27% of the women with borderline tumor had RMI values exceeding 200 and only 37% had CA125 values above the threshold level of 35 U/ml. The ultrasound picture is today probably the best discriminator between benign tumor and ovarian cancer. The diagnostic accuracy of frozen section histology is high for ovarian carcinomas, but the accuracy rate for borderline tumors is reported as relatively low (22), and this was confirmed in our study. It is well known that for a definite diagnosis of borderline ovarian tumor and to rule out tumor invasiveness, a large number of tumor sections are necessary. Due to the restricted time for a patient in the operating theater under general anesthesia, a thorough multiple section analysis is not feasible for the pathologist. Frozen section diagnosis will therefore have a lower sensitivity than routine histology done after tissue fixation (22). In our study, we found a lower accuracy for frozen section histology for mucinous tumors compared to serous tumors, in agreement with other studies (22). This low accuracy rate is probably the reason for the low number of frozen section analyses on ovarian tumors performed at present in our unit. Borderline tumors affect younger patients than invasive epithelial ovarian cancers do. Especially for young patients with early-stage borderline tumor, who wish to preserve their fertility, the laparoscopic approach is preferable. The risk of abdominal postoperative adhesions is considered to be lower after laparoscopy. Conservative surgery of a borderline ovarian tumor with unilateral oophorectomy and omentectomy is considered to be safe, provided the contralateral ovary is macroscopically normal (9). Biopsy from the remaining normal-looking ovary is unnecessary and wedge biopsy could lead to reduced fertility (9). Ovarian resection or cystectomy should be avoided if the cyst looks suspicious on preoperative ultrasound examination, and the surgeon should try to avoid rupture of the cyst and intra-abdominal spilling. Querleu and LeBlanc recommend in every case of a borderline ovarian tumor, when the first operation was performed by laparoscopy, an excision of the trocar sites during the next operation (23). According to Kaern et al., all aneuploid borderline ovarian tumors should be radically operated, including bilateral adnexectomy (24). In our cohort of 107 women operated on for stage I borderline tumors, we had tumor ploidity evaluation performed in 93 patients, revealing 11 aneuploid tumors; the rest were diploid. All our patients with aneuploid borderline ovarian tumors had bilateral adnexectomy performed, except for one patient, who strongly

7 626 E. Ødegaard et al. wished for a pregnancy. These women received no chemotherapy but had a more thorough follow-up than women with diploid borderline tumors. Our material is too small and follow-up too short to conclude that aneuploid tumors must be treated more radically than diploid ones. In conclusion, our data indicate that for stage I borderline tumors, laparoscopic management is feasible if tumor is of moderate size (diameter below 10 cm), gives less postoperative complications and a shorter hospital stay than if managed by laparotomy. The survival seems to be excellent after both types of treatments. Fertility data are limited, and more studies are needed to conclude whether laparoscopy is especially preferable for younger patients wishing to preserve fertility. Limitations of our borderline ovarian tumor study are the lack of randomization, retrospective analysis, and differences in the treatment groups as to tumor size and RMI score. We found that intraoperative tumor rupture is more frequent with the laparoscopic approach, at least for the larger tumors exceeding 10 cm in diameter. Our retrospective study cannot conclude at which precise tumor size cut-off a laparotomy approach is recommended in malignant suspected ovarian tumors. Nevertheless, we believe that our data support that the laparoscopic approach is safe, as long as peroperative tumor spilling is avoided with the use of an endobag. Long-term follow-up of larger patient cohorts is necessary to determine the ultimate clinical outcome, especially after peroperative tumor rupture and spilling. Acknowledgements The work of EØ is supported by research grants from Ulleval University Hospital, and Eastern Norway Regional Health Authority. References 1. Ozols RF, Rubin SC, Thomas G, Robboy S. Epithelial ovarian cancer. In: Hoskins WJ, Perez CA, Young RC, editors. Principles and practice of gynecologic oncology, 2nd edn. Philadelphia, PA: Lippincott-Raven; p Trimble CL, Kosary C, Trimble EL. Long-term survival and patterns of care in women with ovarian tumors of low malignant potential. Gynecol Oncol. 2002;/86:/ Pecorelli S, Odicino F, Maisonneuve P, Creasman W, Shepherd J, Sideri M, et al. Carcinoma of the ovary. J Epidemiol Biostat. 1998;/3:/ Lee KR, Tavassoli FA, Prat J, Dietel M, Gersell DJ, Karseladze AI, et al. Tumours of the ovary and peritoneum. In: Tavassoli FA, Devilee P, editors. Pathology and genetics of tumours of the breast and female genital organs. Lyon: IARC Press; p Fauvet R, Boccara J, Dufournet C, Poncelet C, Darai E. Laparoscopic management of borderline ovarian tumors: results of a French multicenter study. Ann Oncol. 2005;/16:/ Rao GG, Skinner E, Gehrig PA, Duska LR, Coleman RL, Schorge JO. Surgical staging of ovarian low malignant potential tumors. Obstet Gynecol. 2004;/104:/ Morice P, Camatte S, Wicart-Poque F, Atallah D, Rouzier R, Pautier P, et al. Results of conservative management of epithelial malignant and borderline ovarian tumours. Hum Reprod Update. 2003;/9:/ Tingulstad S, Hagen B, Skjeldestad FE, Halvorsen T, Nustad K, Onsrud M. The risk-of-malignancy index to evaluate potential ovarian cancers in local hospitals. Obstet Gynecol. 1999;/93:/ Trope CG, Kristensen G, Makar A. Surgery for borderline tumor of the ovary. Semin Surg Oncol. 2000;/19:/ Quirk JT, Natarajan N. Ovarian cancer incidence in the United States, Gynecol Oncol. 2005;/97:/ Camatte S, Morice P, Atallah D, Thoury A, Pautier P, Lhomme C, et al. Clinical outcome after laparoscopic pure management of borderline ovarian tumors: results of a series of 34 patients. Ann Oncol. 2004;/15:/ Canis M, Mashiach R, Wattiez A, Botchorishvili R, Rabischong B, Jardon K, et al. Frozen section in laparoscopic management of macroscopically suspicious ovarian masses. J Am Assoc Gynecol Laparosc. 2004;/11:/ Maneo A, Vignali M, Chiari S, Colombo A, Mangioni C, Landoni F. Are borderline tumors of the ovary safely treated by laparoscopy? Gynecol Oncol. 2004;/94:/ Vergote I, De BJ, Fyles A, Bertelsen K, Einhorn N, Sevelda P, et al. Prognostic importance of degree of differentiation and cyst rupture in stage I invasive epithelial ovarian carcinoma. Lancet. 2001;/357:/ Sevelda P, Dittrich C, Salzer H. Prognostic value of the rupture of the capsule in stage I epithelial ovarian carcinoma. Gynecol Oncol. 1989;/35:/ Dembo AJ, Davy M, Stenwig AE, Berle EJ, Bush RS, Kjorstad K. Prognostic factors in patients with stage I epithelial ovarian cancer. Obstet Gynecol. 1990;/75:/ Sjovall K, Nilsson B, Einhorn N. Different types of rupture of the tumor capsule and the impact on survival in early ovarian carcinoma. Int J Gynecol Cancer. 1994;/4:/ Berek JS. Ovarian cancer spread: is laparoscopy to blame? Lancet. 1995;/346:/ Querleu D, Papageorgiou T, Lambaudie E, Sonoda Y, Narducci F, LeBlanc E. Laparoscopic restaging of borderline ovarian tumours: results of 30 cases initially presumed as stage IA borderline ovarian tumours. BJOG. 2003;/110:/ Tozzi R, Kohler C, Ferrara A, Schneider A. Laparoscopic treatment of early ovarian cancer: surgical and survival outcomes. Gynecol Oncol. 2004;/93:/ Vaisbuch E, Dgani R, Ben-Arie A, Hagay Z. The role of laparoscopy in ovarian tumors of low malignant potential and earlystage ovarian cancer. Obstet Gynecol Surv. 2005;/60:/ Geomini P, Bremer G, Kruitwagen R, Mol BW. Diagnostic accuracy of frozen section diagnosis of the adnexal mass: a metaanalysis. Gynecol Oncol. 2005;/96:/ Querleu D, LeBlanc E. Laparoscopic surgery for gynaecological oncology. Curr Opin Obstet Gynecol. 2003;/15:/ Kaern J, Trope CG, Abeler VM. A retrospective study of 370 borderline tumors of the ovary treated at the Norwegian Radium Hospital from 1970 to A review of clinicopathologic features and treatment modalities. Cancer. 1993;/ 71:/