Borderline ovarian tumours account for 10–20% of all epithelial ovarian cancer. Historically, standard primary surgery has included borderline ovarian tumours, omentectomy, peritoneal washing and multiple biopsies. As one-third of borderline ovarian tumours are diagnosed in women under the age of 40 years, fertility-sparing treatment has been more frequently used in the past 10 years. Fertility drugs are well tolerated in women with infertility after fertility-sparing surgery. Careful selection of candidates is necessary. Laparoscopic techniques can be used, but should be reserved for oncologic surgeons. This conservative treatment increases the rate of recurrence, albeit with no effect on survival. The pregnancy rate is nearly 50%, and most are achieved spontaneously. These women should be closely followed up. The question is whether this is acceptable from a gynaecologic oncologic point of view. For this reason, we will discuss recently published studies and gynaecologic oncologic concerns about the mode of fertility-sparing surgery and its consequences.
Introduction
Borderline ovarian tumours (BOT) were first described as a separate group in 1929 by Taylor, characterised by histopathologic features and biologic behaviour intermediate between clearly benign and frankly malignant ovarian tumours.
In 1973, this group of tumours was accepted by the International Federation of Gynecology and Obstetrics (FIGO) as carcinoma of low malignant potential, and from 1973 by the World Health Organization as BOT, the term we use today. MostBOT are comparable to benign cysts, but 10–15% may develop a clinical aggressive behaviour, as invasive carcinoma, and have the potential to spread beyond the ovary and have the ability to recur as carcinomas, resulting in poor patient prognosis.
The histologic diagnosis of BOT is based on criteria described by Hart and Norris and detailed by Scully : epithelial cellular proliferation (stratification of the epithelial lining of the papillae, multi-layering of the epithelium, mitotic activity and nuclear atypia) without stromal invasion. The absence of obvious stromal invasion is the principal diagnostic criteria for BOT. The degree of microinvasion ‘accepted’, however, depends on the individual pathologist with problematic inter- and intra-observer reproducibility.
Incidence
Borderline ovarian tumours represent between 10 and 20% of all epithelial ovarian malignancies. At the Norwegian Radium Hospital and other large institutions, the reported prevalence is lower, at 12%, probably reflecting that women with BOT are treated at local hospitals. Few population-based studies on BOT are available. Between 1970 and 1993, a total of 2343 women were diagnosed with BOT in Norway. The age-adjusted incidence rate has increased since 1970 to 4.8 per 100,000 persons per year between 1995 and 2004. Nearly the same trend has been reported in Sweden.
In a recent review of 6362 women with BOT by Trillsch et al., 78.9% of women had stage I disease and few had abdominal spread, whereas stage IV disease represented an exception. These figures are in agreement with a large retrospective Norwegian study of 370 women with BOT treated at the Norwegian Radium Hospital between 1970 and 1982. In the latter study, 311 women (84%) were diagnosed with stage I disease, 20 women (6%) with stage II, and 39 women (10%) at stage III. None had stage IV disease. Median age at diagnosis was 40 years (27% were younger than 40 years of age) compared with about 60 years for women with invasive carcinoma, and the incidence rate increased with age up to 45–49 years, after which the rate stabilised. Between 1970 and 1993 in Norway, 93% of women were diagnosed with localised tumours. The proportion of tumours with distant metastases increased from 2.7% between 1970 and 1973 to 6.6% between 1989 and 1993.
Incidence
Borderline ovarian tumours represent between 10 and 20% of all epithelial ovarian malignancies. At the Norwegian Radium Hospital and other large institutions, the reported prevalence is lower, at 12%, probably reflecting that women with BOT are treated at local hospitals. Few population-based studies on BOT are available. Between 1970 and 1993, a total of 2343 women were diagnosed with BOT in Norway. The age-adjusted incidence rate has increased since 1970 to 4.8 per 100,000 persons per year between 1995 and 2004. Nearly the same trend has been reported in Sweden.
In a recent review of 6362 women with BOT by Trillsch et al., 78.9% of women had stage I disease and few had abdominal spread, whereas stage IV disease represented an exception. These figures are in agreement with a large retrospective Norwegian study of 370 women with BOT treated at the Norwegian Radium Hospital between 1970 and 1982. In the latter study, 311 women (84%) were diagnosed with stage I disease, 20 women (6%) with stage II, and 39 women (10%) at stage III. None had stage IV disease. Median age at diagnosis was 40 years (27% were younger than 40 years of age) compared with about 60 years for women with invasive carcinoma, and the incidence rate increased with age up to 45–49 years, after which the rate stabilised. Between 1970 and 1993 in Norway, 93% of women were diagnosed with localised tumours. The proportion of tumours with distant metastases increased from 2.7% between 1970 and 1973 to 6.6% between 1989 and 1993.
Pathogenesis
Recently, a new theory has been developed describing a subset of serous ovarian cystadenomas that evolve through serous BOT to low-grade epithelial ovarian cancer ( Fig. 1 ). This ‘low-grade’ pathway involves mutations in the B-raf and K-ras signalling pathway. The pathway involves serous BOT as a precursor mimicking the adenocarcinoma sequence in colorectal cancer, in which carcinoma evolves through a continuum of histological precursor lesions. Only 2% of all serous BOT progress to carcinoma via the ‘low-grade’ pathway.
This is in contrary to the ‘high-grade’ pathway that involves frequent TP53 mutations and very seldom B-raf and K-ras mutations. Most serous ovarian carcinomas belong to the typical high-grade pathway, with no known precursor. TP53 mutations are most often absent in typical serous BOT and micropapillary serous BOT compared with 88% prevalence in invasive serous carcinomas.
The sequence of malignant transformation from benign mucinous tumours to carcinoma represents transitional stages of mucinous carcinogenesis. Three types of ras oncogenes exist (K, N and H) and mucinous BOT have a higher frequency of K-ras mutation than that of mucinous cystadenoma, but a lower rate than that of mucinous carcinoma. Foci suggesting in-situ malignant changes are often identified in mucinous BOT associated with invasion, and the transition may be identified microscopically in some cases. It is for the moment unclear whether BRCA1 and BRCA2 mutations increase the risk of BOT.
Serous borderline tumours
Serous borderline tumours are bilateral in about one-third of cases. Extra-ovarian spread as peritoneal implants is frequent (35%). Most implants are non-invasive, but invasive implants are found in 20–25% of cases. Bilaterality and non-invasive implants do not predict worse outcome compared with the presence of invasive implants. Invasive implants may progress to invasive carcinoma, whereas most peritoneal implants will remain stable or regress after removal of the main ovarian tumour. Women with serous BOT without invasive implants have 10-year survival of 95% compared with 60–70% for women with invasive implants. Women with invasive implants develop progressive disease in about 30% of cases, whereas only about 2% of women without invasive implants will progress. Controversy exists about whether tumours with invasive implants are more likely to have micropapillary features within the primary BOT at diagnosis.
Mucinous borderline tumours
Mucinous borderline tumours are histologically classified as intestinal (85% of mucinous BOT) or endocervical (15% of mucinous BOT) (former ‘Müllerian’). The ovarian tumours of the intestinal subtype can be large, and are nearly always unilateral. In case of bilaterality, the woman should be further examined for a primary intestinal tumour. The endocervical subtype might be bilateral and associated with endometriosis and mixed-BOT. Both subtypes can present with intra-epithelial carcinoma, micro-invasion with less than 10 mm Extraovarian spread is infrequent (10–15%) and nearly always as pseudomyxoma peritonei. The condition may occur in any type of intra-abdominal mucinous neoplasm but occurs most often in association with mucinous BOT. Most of these represent dissemination from a mucinous appendix tumour. An appendectomy should be carried out at primary surgery for intestinal M-BOT. All M-BOT might recur as invasive adenocarcinoma, especially when only ovarian cystectomy rather than salpingo-oophorectomy has been carried out. This might be explained by missing an invasive component owing to inadequate histopathologic evaluation of these large tumours. Therefore, unilateral salpingo-oophorectomy is recommended in mucinous BOT.
Prognostic factors
As BOT in general has an excellent survival, it might be difficult to identify variables that improve survival. As many women with BOT are of reproductive age at diagnosis, more conservative treatment preserving the childbearing ability is desirable. International Federation of Gynecology and Obstetrics stage is the strongest prognostic factor for recurrence and survival for BOT, as well as for invasive ovarian cancers. Micropapillary histology was reported as an additional risk factor for serous BOT, but this issue is controversial, as poor prognosis is only seen if this histology is associated with invasive implants. Tumours with evidence of stromal invasion of less than 10 mm 2 behave clinically as BOT and are today classified as BOT. In the retrospective study from the Norwegian Radium Hospital, 370 BOT were treated between 1970 and 1982 and the follow up was complete, with a median follow-up time of 152 months (range 6–354 months). By multivariate analysis, the only three independent prognostic factors for disease-free and long-term survival were FIGO stage ( P < 0.0001), histologic type ( P < 0.05) and age ( P < 0.005). In another study of the same women, including DNA ploidy, the women could be divided into risk groups. The low-risk group (100% disease-free survival) was characterised by stage I disease, diploid serous BOT or mucinous BOT and age less than 40 years. The high-risk group (75% or higher risk of dying of disease) had aneuploid serous BOT or mucinous BOT, stage II-III disease and age older than 70 years. Deoxyribonucleic acid ploidy was the strongest prognostic factor in BOT in the study by Kaern et al., but this has unfortunately not been reproduced by others. Lymph-node involvement could not convincingly be confirmed to be an independent risk factor.
Symptoms
Limited information exists about symptom registration by women with BOT at diagnosis. Most women may be asymptomatic and a pelvic mass discovered at routine pelvic examination. Symptoms are less common compared with symptoms registered for women with invasive ovarian cancer, although 75% of women with BOT may have at lease one symptom as abdominal pain or discomfort, bowel irregularity and persisting fatigue or weight loss.
Standard treatment
The diagnosis of BOT cannot be determined before surgery. Intraoperative frozen section diagnosis of BOT is often difficult, even for experienced pathologists. Frozen sections may be useful for discrimination between BOT and epithelial ovarian cancer and benign tumour (over-diagnosis less than 10%), but not for discrimination between BOT and epithelial ovarian cancer (under-diagnosis in 25–30% of women). Less than 50% of women with BOT will have a complete surgical staging, this without obvious influence on survival.
The standard guidelines for primary surgery in BOT are similar to these for invasive ovarian carcinomas; removal of all macroscopic disease and proper surgical staging: hysterectomy with bilateral salpingo-oophorectomy, multiple peritoneal biopsies and peritoneal washing with cytology. For mucinous BOT, appendectomy should also be carried out. Lymph-node sampling has not been part of the standard procedure. A recently published German study reported systematic lymph-node dissection in 18.9% of women with BOT. Involvement of lymph nodes, even with upstage, do not influence recurrence or survival rate. Lymph-node involvement is rarely seen in M-BOT. Lymphadenectomy can be omitted even for stage II and III disease, as there is no difference in the recurrence or survival rate ( Fig. 2 ).
Restaging procedures
Despite the fact that experts have recommended comprehensive surgical staging for many years, most women referred to university or comprehensive cancer centres after primary surgery have incomplete surgery. Another common event is that malignancy is not suspected at the initial operation, and therefore the abdomen is not properly explored. Should this woman undergo re-operation? Snider et al. upstaged five out of 27 women with stage I disease. None of the 12 women with mucinous BOT were upstaged, whereas five out of 13 women with serous BOT were upstaged. Mucinous BOT grossly confined to one ovary are probably unlikely to be upstaged at a restaging procedure. If evidence of extra-ovarian disease exists, complete surgical staging and appendectomy are indicated. Mucinous BOT with abdominal spread are in 50% claimed to arise in the appendix or simultaneously in one or both ovaries. For serous BOT, it is reasonable to do a second operation, as lymph-node involvement has been reported in about 20% in apparent stage I tumours, with even higher incidence in more advanced stages. At the Norwegian Radium Hospital, we recommend restaging if the women are insufficiently staged at primary operation, in agreement with Cadron et al. and Trimble and Trimble. We also believe that the decision to restage must be individualised, taking into consideration the adequacy of the previous abdominal exploration, the tumour subtype, and potential treatment of residual disease, as well as the level of concern for both the woman and the clinician, even if there is little proof that restaging operations have an effect on survival.
Laparoscopy
Laparoscopic treatment has become an attractive approach for benign ovarian tumours (pelvic mass). The development of the laparoscopic techniques, with the use of endo-bag, gives less risk of spillage during surgery and less port-side metastases. For conservatively treated women wishing to preserve fertility, this is a more favourable approach, and has short postoperative recovery, fewer adhesions and improved cosmetic results. Cyst rupture and incomplete staging occur significantly more frequently by laparoscopy compared with laparotomy (33.9% v 12.4%), but do not influence survival. The laparoscopic treatment may result in higher recurrence rate, which must be discussed with the woman. The woman should agree to close follow up and repeated surgery. Other concerns are port metastases and improper staging.
Postoperative treatment of borderline ovarian tumours
Today, there is no proven benefit from adjuvant therapy (chemotherapy or radiotherapy) even in advanced stage disease and in the presence of invasive implants. No randomised-controlled studies have evaluated the efficacy of chemotherapy in advanced stage BOT with residual disease. Cisplatin-based therapy in serous BOT with residual disease have shown high response rates but modest effect on long-term survival. In case of recurrences where transformation to an invasive histology has occurred, consideration of platinum-taxane-based chemotherapy is reasonable. The molecular changes transforming BOT to invasive ovarian carcinoma may be clarified in the future with the possibility of blocking the different ‘steps’ in the carcinogenesis. Today, we know that both serous BOT and low-grade serous carcinoma frequently present with B-raf and K-ras mutations.
Survival
Overall survival for women with BOT is excellent: 90–100% in most reports depending on age at diagnosis, FIGO stage and histologic type. Trillsch et al. reported a 5-year overall survival rate of almost 100% in early stage disease (FIGO I-II) and between 86% and 92% in more advanced disease (FIGO III-IV). In a Norwegian population-based study, the age-adjusted 1-, 3-, 5- and 10-year relative survival rates of women with mucinous BOT were 98%, 97%, 97%, and 95%, respectively, and for women with serous BOT 96%, 93%, 90%, and 90%, respectively. The relative risk of dying increased with higher age at diagnosis. The 5-year relative survival for women aged between 0 and 44 years was 99% compared with 85% for women aged 75–89 years. Nearly the same figures are reported by Sherman et al., with better survival for younger women and slightly better survival for serous tumours compared with mucinous tumours, although mucinous tumours in advanced stage had the poorest long-term survival (85.5%).
Seidman and Kurman summarised 97 reports, including a total of 4129 women with serous BOT, and showed a disease-specific survival rate of 99% for stage I disease, and 95.3% for stage III disease (mean follow up of 7 years). The 5-year survival was 98% for women with serous BOT with non-invasive implants and 33% for those with invasive implants.
Conservative surgery
The median age of diagnosis for women with BOT is 40 years, and nearly one-third are younger than 40 years. Thus, preservation of childbearing potential is an important issue for many of these women. To select women for fertility-sparing surgery without risk of reduced long-term survival is a challenge for the gynaecologic oncologist ( Fig. 2 ). Conservative surgery is defined as complete staging in which the uterus, and at least a part of one ovary, are preserved ( Fig. 2 ).
Although the standard treatment for all women with BOT is at least bilateral salpingo-oophorectomy, many young women with stage Ia tumours who have not completed childbearing can be safely treated with unilateral salpingo-oophorectomy after comprehensive surgical staging, thereby preserving the fertility potential. At least as important as the fertility issue, however, is whether we can reduce the morbidity caused by radical surgery and whether a more conservative approach is a safe alternative regarding cancer prognosis. Recent studies from the Norwegian Radium Hospital and Gynecologic Oncology Group have shown that preservation of reproductive organs is feasible. Relapse rates after bilateral salpingo-oophorectomy range between 0% and 20%. This rate varies between 12% and 58% for cystectomy and between 2.5% and 5.7% for radical surgery. Relapses can also occur late, and recurrences as late as 39 years after initial treatment have been reported. Predictors of relapse are resection margins containing tumour cells. The frequency of persistence or recurrence in 35 patients with serous-BOT treated by unilateral cystectomy, bilateral cystectomy or unilateral cystectomy with contralateral oophorectomy or salpingo-oophorectomy was retrospectively investigated by Lim-Tan et al. Conservative surgical treatment was carried out either because the women were young and wanted to preserve their fertility or because the nature of the tumour was not determined at the time of surgery. Thirty-three women had stage I disease (19 stage Ia, 10 stage Ib and four stage Ic), and two had stage III disease. Although 60% of the women had additional and in some cases definitive operations within relative short periods after initial cystectomy or cystectomies, 21 had conservation of ovarian tissue and were followed on average for 7.5 years. All women with stage I disease were alive without evidence of disease after surgical treatment alone. The presence of persistent or recurrent disease correlated with multifocality and involvement of resection margins. Multifocality may be a strong predictor of failure of cystectomy to control the disease. No recurrence of disease was seen in ovaries from which a single cyst had been removed with negative resection margins. Extensive sampling of the resection margins of ovarian cysts is important.
For many women with BOT, fertility is an important issue. It is, therefore, essential to discuss with these young women the advantages and disadvantages between conservative and radical surgery. Conservative surgery preserves fertility and hormonal function, but unfortunately with high risk of recurrence.
In women with obvious stage I disease, unilateral salpingo-oophorectomy or cystectomy has from previous studies been shown to be safe. This has also been confirmed for women with advanced-stage disease on the assumption of close follow up.
Even for women with advanced disease, conservative surgery might be considered after informed consent. The recurrence rate is high, and invasive and even non-invasive implants may recur up to 15% as invasive OC without negative influence on survival.
Stay updated, free articles. Join our Telegram channel
Full access? Get Clinical Tree
