Vol. 14 No 1 | Autumn 2012
Epithelial ovarian cancer
Dr Deborah Neesham

This article is 12 years old and may no longer reflect current clinical practice.

Ovarian cancer is the ninth most common cancer diagnosed in Australian women, with one in 80 being diagnosed by age 85, and is the seventh most common cause of cancer death. It is the most lethal gynaecological cancer, with around 1300 women diagnosed each year and around 850 deaths.1

In over 90 per cent of women with localised disease, surgery alone is curative, but unfortunately the majority of women have advanced disease at diagnosis. Although there has been an apparent improvement in five-year survival in the last 20 years, from 33 to 40 per cent, this has probably been the result of improved long-term survival with disease rather than an increase in cure rate. In the absence of any useful methods to diagnose the disease at an earlier stage, prevention should be a primary aim.

Adjuvant chemotherapy with platinum and paclitaxel remains the standard of care following surgery, but newer agents that target specific biological markers are being developed and, with advances in gene profiling, individualisation of treatment is the hope of the future.


    1. The oral contraceptive pill (OCP) reduces the risk of ovarian cancer in the general population by up to 80 per cent with long-term use. The OCP also appears to reduce the risk of ovarian cancer in gene mutation carriers, such as BRCA1/2 (in case-controlled studies only), with 40 per cent reduction with long-term use and no apparent increase in breast cancer risk.2
    2. Hysterectomy reduces the relative risk by 50 per cent.
    3. Tubal ligation reduces the relative risk by 50 per cent.
    4. Bilateral salpingo-oophorectomy (BSO) reduces the risk by 95 per cent. The residual risk lies in developing primary peritoneal cancer.


Screening is not effective in the diagnosis of precancerous lesions or early-stage ovarian cancer based on available information, although results of a large randomised trial of ultrasound versus CA-125 versus expectant management in over 200 000 postmenopausal women in the UK is expected in the next 12 months (UKTOCS). Although CA-125 is usually elevated (over 90 per cent) in advanced epithelial ovarian malignancies, it is elevated in only 50 per cent of early-stage cancers of the ovary, which limits its use in screening as well as diagnosis.

Transvaginal ultrasound is the other investigation commonly used as a screening tool, but at present lacks both sensitivity and specificity.

As a result of current knowledge of available screening tests, the 2011 National Breast and Ovarian Cancer Centre guidelines, endorsed by RANZCOG, do not recommend screening in the general population or in women at increased risk due to family history or a known gene mutation.3

There is ongoing promising research assessing multiple markers that may be indicative of early ovarian cancer, but none of these, to date, have undergone full evaluation in a population setting.


CA-125 measurement is helpful in investigating a postmenopausal woman with a pelvic mass as, in this situation, an elevated CA- 125 indicates malignancy in around 90 per cent of cases, whereas pre-menopausally, endometriosis, adenomyosis, menstruation, fibroids, ovulation and pregnancy can all raise the CA-125 level, limiting its interpretation.

Carcinoembryonic antigen (CEA), CA 19-9 and CA 15-3 may be elevated in mucinous tumours, but these account for less than five per cent of epithelial malignancies and if elevated are more likely to indicate a gastro-intestinal, pancreatic or breast primary. Serum inhibin is also often elevated in mucinous tumours, but is most helpful in the management of granulosa cell tumours. Currently, the timeframe for obtaining the results of inhibin assays makes this relatively unhelpful at the time of primary diagnosis.

In most situations, pelvic masses still present a diagnostic dilemma. Around 20 per cent of ovarian cancers occur before menopause and differentiating between functional, benign and malignant masses remains problematic, particularly in the setting of endometriosis in which clear cell and endometrioid tumours of the ovary may arise. Ultrasound, in the hands of trained gynaecologists, can help clarify suspicious features, particularly the complexity of a mass and solid areas with low-resistance blood flow (indicating neovascularisation), nodularity in the pouch of Douglas and free fluid indicative of ascites. Newer magnetic resonance imaging (MRI) technology using weighted diffusion images, in experienced hands, can differentiate endometriosis from malignancy, increasing the specificity of imaging and reducing the intervention for benign disease4, but differentiating borderline from malignant tumours is not reliably possible and frozen section remains the best method of planning intra-operative management. However, frozen section may still under-call the situation, particularly in the setting of very large complex masses and this must be discussed pre-operatively with the woman. The most important aspect of managing these masses is not to convert a stage IA cancer into an advanced cancer. If masses are to be removed they should be removed intact or in a retrieval bag, if at laparoscopy. Morcellation of solid masses intraperitoneally should be avoided.



If an ovarian mass clinically confined to the ovary is found to be malignant at frozen section or final histopathology, then full surgical staging (including para-aortic lymph node dissection, omental sampling and peritoneal sampling with peritoneal washings) is required for both treatment planning and prognosis. If there is tumour rupture or surface involvement, adjuvant chemotherapy is usually recommended. Fertility-sparing surgery may be considered in women wanting to conceive, but this should only occur after extensive counselling and only for stage IA cancer and stage IC disease with favourable histological features.

Advanced ovarian cancer requires cytoreduction, with the amount of residual disease being one of the strongest prognostic factors. Leaving no macroscopic residual disease is now considered optimal cytoreduction. If there is no macroscopic residual disease at the completion of surgery the median progression-free survival (PFS) is around 100 months, compared to the historical mean of 18 months.5

To achieve optimal cytoreduction to no macroscopic residual, more aggressive surgical procedures are commonly required, such as rectosigmoid resection, splenectomy and liver mobilisation to aid resection of hepatic and diaphragmatic disease, including pleural resection. Patient selection is important in terms of identifying candidates for these morbid procedures, as miliary small bowel or mesenteric disease, porta hepatis disease, high para-aortic nodal disease or extensive distant metastases may preclude optimal cytoreduction and render this surgery futile. Extensive ascites (over 1000ml) appears to be a predictor of inability to cytoreduce to no macro residual. In these situations neo-adjuvant chemotherapy for stage IIIC and stage IV disease is often preferred with the aim of maximal cytoreduction after three cycles of chemotherapy. A recent randomised controlled trial in Europe showed no adverse effect on disease-free survival or mortality and a reduction in surgical morbidity utilising neoadjuvant chemotherapy in this group of patients.6


Standard chemotherapy for advanced epithelial ovarian cancer remains six cycles of intravenous (IV) combination carboplatin and paclitaxel given every three weeks. This treatment has a response rate of 80 per cent and a PFS of 18/12. New modes of administration of chemotherapy such as intraperitoneal (IP) administration have been investigated and preliminary trials have shown an improvement in PFS but with increased toxicity. Trials are ongoing to assess whether less morbid IP regimens are more effective than our current mode of intravenous administration.

Changing the schedule of administration of chemotherapy may also alter its effectiveness. A large Japanese study showed that a regimen of paclitaxel given weekly was more effective than three-weekly administration, increasing PFS from 17 to 28 months, but unfortunately also was more myelosuppressive, probably due to a higher overall dose of paclitaxel than with the standard chemotherapy regimen.7 This regimen is currently being assessed in Western populations.

New biological agents

Newer agents, such as angiogenesis inhibitors (for example, bevacizumab), target the tumour micro-environment, not the tumour itself, and in randomised controlled trials have shown activity in the adjuvant and recurrent setting, extending PFS by four to six months with prolonged use.8 Bevacizumab has also shown activity as a single agent in the treatment of bulky recurrent disease and extensive ascites, but has to date demonstrated no improvement in overall survival. This agent is often associated with side effects, such as hypertension, diarrhoea, abdominal pain, fatigue and asthenia, with serious side effects including bowel perforation, haemorrhage and arterial thrombo-embolism. It is also expensive, costing around $10 000 per month. There is no biomarker currently available to determine which patients would benefit from its use.

Poly (ATP-ribose) polymerase inhibitors (PARPIs) (for example, olaparib, iniparib) are also proving useful in the treatment of BRCA mutation related high-grade serous ovarian cancer (around 20 per cent of these high-grade serous tumours) and also prolong PFS by around four months in advanced disease.9

The real hope for individualisation of treatment lies in new technology that allows the measurement of expression of thousands of genes, with the ability to demonstrate their clinical phenotypes. Differences in intrinsic, as opposed to acquired, chemoresistance of ovarian tumour cell lines has already been demonstrated. Several studies have tried to identify gene expression signatures correlating with survival and relapse and the new challenge is to develop gene expression profiles that could be used to develop early detection or screening tests and determine individual treatment plans.


Ovarian cancer remains one of the deadliest female cancers, with over two-thirds of women having advanced disease when diagnosed. Disease prevention is the ideal, with the OCP providing significant risk reduction as do pregnancy, tubal ligation and hysterectomy.

Subtypes of ovarian cancer

It has become increasingly obvious that there is a great variation in ovarian cancer and its behaviour. Ovarian cancer can no longer be considered to be one disease. Molecular biology has improved our understanding of the traditional subtypes.

  1. Serous ovarian cancers (80 per cent) can be separated into two types:
    1. Low-grade Type 1 with mutations in K-Ras, BRAF, PTEN, CTNNB-1, b-catenin that probably arise from ovarian cortical inclusion cysts; and
    2. High-grade Type 2 with mutations in p-53. There is increasing evidence that the fallopian tube may be the origin of many of these high-grade serous tumours.10 The ovarian epithelial surface accounts for only one per cent of the total ovarian mass despite over 90 per cent of primary ovarian tumours being epithelial in origin. P-53 signatures can be found in benign appearing tubal epithelium and tubal intraepithelial cancer has been found in five to ten per cent of women with serous ovarian cancers and in BSO specimens of women having prophylactic surgery who are at high risk of ovarian cancer. BRCA mutations have been found in 18 per cent of women diagnosed with high-grade serous cancer, regardless of their family history. Prophylactic tubal removal in women who have completed childbearing and who are undergoing pelvic surgery for benign reasons should be seriously considered.
  2. Endometrioid ovarian cancers (ten per cent) are associated with endometriosis. Risk is reduced with increased parity, OCP use, breastfeeding and tubal ligation. Steroid receptors are commonly positive and response to progestogens has been described.
  3. Clear cell ovarian cancers (five per cent) are associated with endometriosis and obesity. Reduced risk occurs with OCP use and increased parity. Mutations in ARID1 and PIK3CA are common. Of these, 40–60 per cent are diagnosed with early stage disease, but recurrence is common with a median PFS of 12/12. If disease is advanced at diagnosis, 50 per cent progress on first line standard chemotherapy with an overall complete response rate of only 11 per cent.
  4. Mucinous ovarian cancers (less than five per cent) are rare as primary ovarian tumours and gastro-intestinal metastasis must always be excluded. They are associated with smoking and obesity. There is now some evidence for progression from benign to borderline to invasive disease. They uncommonly metastasise to lymph nodes.11


In women at high risk of developing ovarian cancer, screening is currently ineffective and prophylactic surgery with BSO is the only effective risk-reducing mechanism available, but this must be balanced by the morbidity of surgery and surgical menopause. Prophylactic removal of the fallopian tubes should be discussed with all women who have completed their families and are undergoing pelvic surgery for benign disease.

First-line treatment of epithelial ovarian cancer has not changed in principle, with the aim of optimising cytoreduction to no macroscopic residual. Increased utilisation of more extensive surgery, including liver, spleen, diaphragmatic and pelvic rectosigmoid resection, affords dramatically improved outcomes for the women in whom this is possible. Adjuvant chemotherapy with platinum and paclitaxel remains the standard of care, but new administration routes and scheduling are under investigation. Newer biological agents targeting angiogenesis and other molecular targets delay tumour growth, and many new agents are under investigation. However, the challenge for ovarian cancer lies in the development of effective screening or early detection tests and also individualisation of treatment, particularly as more information suggests that the subtypes of epithelial ovarian cancer are heterogenous in their gene expression and thus in their response to treatment.


I wish to thank Prof Michael Quinn who proofread and edited this article.


  1. Australian Institute of Health and Welfare & Australasian Association of Cancer Registries 2010. Cancer in Australia: an overview, 2010.Cancer series no. 60. Cat. no. CAN 56. Canberra: AIHW.
  2. Iodice S, Barile M, Rotsmensz et al. Oral contraceptive use and breast or ovarian cancer risk in BRCA 1/2 carriers: A meta-analysis. EJC 2010.Vol26:2275-2284.
  3. Population screening and early detection of ovarian cancer in asymptomatic women – NBOCC Position Statement 2009.
  4. Iyer VR, Lee SI. MRI, CT and PET/CT for Ovarian Cancer Detection andAdnexal Lesion Characterisation. AJR 2010;Vol194no2:311-321.
  5. du Bois A, Reuss A, Pujade-Lauraine E et al. Role of surgical outcome as prognostic factor in advanced epithelial ovarian cancer: a combined exploratory analysis of 3 prospectively randomized phase 3 multicenter trials: by the Arbeitsgemeinschaft Gynaekologische OnkologieStudiengruppe Ovarialkarzinom (AGO-OVAR) and the Grouped’Investigateurs Nationaux Pour les Etudes des Cancers de l’Ovaire(GINECO) Cancer 2009;115:1234–1244.
  6. Vergote I, Trope CG, Amant F, et al. Neoadjuvant Chemotherapy or Primary Surgery in Stage 3C or 4 Ovarian Cancer. N Engl J Med2010;363:943-953.
  7. Katsumata N, Yasuda M, Takahashi F et al. Dose-dense paclitaxel once a week in combination with carboplatin every 3 weeks for advanced ovarian cancer: a phase 3, open-label, randomised controlled trial.Lancet 2009; 374(9698):1331-8.
  8. Burger RA, Brady MF, Rhee J, Sovak MA, Nguyen H, Bookman MA.Independent radiologic review of GOG 218, a phase 3 trial of bevacizumab(BEV) in the primary treatment of advanced epithelial ovarian cancer (EOC), primary peritoneal (PPC), or fallopian tube cancer (FTC). J Clin Oncol 2011;29(suppl): abstract 5023.9
  9. Ledermann JA, Harter P, Gourley C et al. Phase 2 randomised placebo-controlled study of olaparib (AZD2281) in patients with platinum-sensitive relapsed serous ovarian cancer (PSR SOC). J Clin Oncol2011;29(suppl):abstr 5003.
  10. Cancer Genome Atlas Research Network. Integrated genomic analyses of ovarian carcinoma. Nature 2011;474:609-18.
  11. Bell DA. Origins and Molecular Pathology of Ovarian Cancer. ModernPathology 2005. S19-32.

Leave a Reply

Your email address will not be published. Required fields are marked *