Skip to main content
Public beta This website is in public beta – please give your feedback.

Classification of ovarian cancer

90% of ovarian malignancies derive from the epithelial cells covering the outer surface of the ovary. These epithelial ovarian tumours can be further subdivided into:

  • serous ovarian carcinoma (the most common subtype at around 75%);
  • mucinous ovarian carcinoma;
  • endometrioid ovarian carcinoma; and
  • clear cell carcinoma

Rarely, malignancies can also develop in the epithelial cells covering the peritoneum (primary peritoneal cancer) and fallopian tubes. These resemble epithelial ovarian tumours in their origin, progression and treatment.

The remaining 10% of ovarian tumours mainly comprise germ cell tumours arising from the reproductive cells, and sex-cord stromal tumours arising from connective tissues.

Epidemiology of ovarian cancer

Most cases of ovarian cancer occur in post-menopausal women, with incidence rates climbing steadily from age 40, and a peak incidence rate in ages 75–79 years. Inherited ovarian cancer is more likely to occur at a younger age than sporadic cancer.

Ovarian germ cell tumours, in contrast, usually affect young women, with a mean age of presentation of 19 years.

Genetics of ovarian cancer

Somatic (tumour) genetics

Most cases of ovarian cancer are caused by somatic (tumour) mutations acquired during an individual’s lifetime, affecting genes that control cell growth and division or DNA repair mechanisms.

Somatic mutations are those that have arisen in the tumour, and are not present constitutionally; they are therefore not present in the germline and cannot be passed on to offspring.

  • More than half of high-grade serous ovarian cancers (HGSOC) have a disruption on the homologous recombination (HR) DNA repair pathway, by mutation or epigenetic silencing.
  • Tumours demonstrating HR repair deficiency usually have a characteristic pattern of mutations, known as a ‘mutational signature’ (in this case, mutational signature 3) in the tumour genome sequence.
  • HGSOC usually show a high degree of chromosomal instability and copy number variations.
  • Ovarian tumours can be stratified by their HR-deficient or proficient status (such as the Myriad HRD test). Improved survival with PARP inhibitor therapy has been seen in all HR groups in most studies and, therefore, tests of HR status are not required to determine eligibility for PARP inhibitors in most clinical indications. One exception is determining eligibility for olaparib plus bevacizumab maintenance treatment of high-grade advanced ovarian cancer. Eligibility criteria for accessing this treatment regimen from the Cancer Drugs Fund are based on the PAOLA-1 clinical trial and therefore require a complete or partial response after first-line platinum-based chemotherapy plus bevacizumab, and that the cancer is associated with HR deficiency.
  • Most Genomic Laboratory Hubs are currently submitting samples for the Myriad HRD test, but NHS mutational signature detection tests are in development.

Constitutional (germline) genetics

In contrast to the sporadic cases, a significant number of cases of ovarian cancer (15%–25%) have an underlying inherited genetic cause. Cases of ovarian cancer may therefore cluster in families. Often it is not possible to identify an inherited genetic cause. In some families, however, a high-risk, high-penetrance ovarian cancer predisposition gene variant may be identified.

  • The majority of these variants (65%–85%) are in the BRCA1 and BRCA2 genes, and are associated with a high lifetime risk of developing ovarian cancer (36%–53% for BRCA1, 11%–25% for BRCA2; compared with a 1.6% risk in the general population). BRCA1 and BRCA2 are key mediators of DNA repair by homologous recombination.
  • Constitutional variants in other genes involved in homologous recombination repair are also implicated in a small percentage of cases of hereditary ovarian cancer, including RAD51C/RAD51D, PALB2 and BRIP1.
  • A further 10%–15% of cases of hereditary ovarian cancer are caused by variants in the genes involved in DNA mismatch repair (MMR), causing predisposition to a range of cancers, known as Lynch syndrome. Ovarian cancers associated with Lynch syndrome are predominantly endometrioid or clear cell carcinomas. Serous cancers are unlikely to be caused by Lynch syndrome.
  • The extremely rare and aggressive small cell carcinoma of the ovary, hypercalcaemic type, is a malignant rhabdoid tumour and is associated with somatic or constitutional mutations in the SMARCA4 gene.
  • Many non-epithelial ovarian cancers are also associated with hereditary cancer predisposition syndromes: Sertoli-Leydig cancers and gynandroblastomas are strongly associated with DICER1 predisposition syndrome, and ovarian sex cord tumours with annular tubules with Peutz-Jeghers syndrome. Ovarian gonadoblastomas should prompt consideration of an underlying sex chromosome disorder. Adult granulosa cell tumours of the ovary are almost exclusively caused by somatic events in FOXL2.

Management implications of genomic testing

  • If a BRCA1 or BRCA2 variant (either constitutional or somatic) is identified in an individual with ovarian cancer, it may have implications for the choice of treatment for their current cancer. For more on this, see the Knowledge Hub article on PARP inhibitors.
  • If a constitutional variant in an ovarian cancer predisposition gene variant is identified, there are also implications for management of the patient’s own future cancer risk and that of their relatives. The cancer risk associated with variants in currently known ovarian cancer susceptibility genes is inherited as an autosomal dominant trait.


For clinicians


Tagged: Constitutional mutations, Somatic mutations, Ovarian cancer

↑ Back to top
  • Last reviewed: 03/05/2022
  • Next review due: 03/05/2024
  • Authors: Dr Amy Frost
  • Reviewers: Dr Ellen Copson, Dr Claire Green, Dr Helen Hockings, Dr Terri McVeigh