Article Text

Download PDFPDF

MSH2 is the very young onset ovarian cancer predisposition gene, not BRCA1
  1. Nicola Flaum1,2,
  2. Emma J Crosbie3,4,
  3. Emma Roisin Woodward5,
  4. Fiona Lalloo5,
  5. Robert Morgan6,
  6. Neil Ryan7,
  7. D Gareth Evans1,2,6,8,9
  1. 1Division of Evolution, Infection and Genomics, School of Biological Sciences, Faculty of Biology, Medicine and Health, University of Manchester, Manchester Academic Health Science Centre, Manchester, UK
  2. 2North West Genomics Laboratory Hub, Manchester Centre for Genomic Medicine, St Mary’s Hospital, Manchester University NHS Foundation Trust, Manchester Academic Health Science Centre, Manchester, UK
  3. 3Division of Cancer Sciences, School of Medical Sciences, Faculty of Biology, Medicine and Health, University of Manchester, Manchester, UK
  4. 4Division of Gynaecology, St Mary’s Hospital, Manchester University NHS Foundation Trust, Manchester, UK
  5. 5Clinical Genetics Service, Manchester Centre for Genomic Medicine, St Mary’s Hospital, Manchester University NHS Foundation Trust, Manchester, UK
  6. 6The Christie NHS Foundation Trust, Manchester, UK
  7. 7The Academic Women’s Health Unit, Translational Health Sciences, Bristol Medical School, University of Bristol, Bristol, UK
  8. 8Prevention Breast Cancer Centre and Nightingale Breast Screening Centre, University Hospital of South Manchester, Manchester, UK
  9. 9Manchester Breast Centre, Manchester Cancer Research Centre, University of Manchester, Manchester, UK
  1. Correspondence to Dr Nicola Flaum; nicola.flaum{at}

Statistics from

Request Permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.

Referenced paragraph

The genetic cause of very young onset ovarian cancer (VYOC), diagnosed under 30 years of age, is unclear.1 The histology and underlying genetics in VYOC is significantly different from the overall epithelial ovarian cancer (EOC) population; we aimed to explore this in VYOC cases known to the North-West of England. We found mismatch repair genes to be the most commonly affected in VYOCs, especially MSH2. The cumulative likelihood of an EOC in MSH2 heterozygotes is >2% by age 35, with this likelihood still below 0.5% for BRCA1 and rare for BRCA2.2


The inherited landscape of epithelial ovarian cancer (EOC) is well established with contributions from homologous recombination deficiency (HRD) genes, particularly BRCA1 and BRCA2, and mismatch repair deficiency (MMRD) genes MSH2, MLH1, MSH6 and PMS2.1 High-grade serous ovarian cancer (HGSOC) is associated with HRD, accounting for up to 23% of HGSOC.2 Approximately 3% of EOC cases occur in <30 years of age, described as very young onset ovarian cancer (VYOC).1 The pathology in VYOC differs from overall EOC; a study of 114 VYOC cases found only 28% were serous, while 59% had mucinous pathology.3 Among the 101 tested cases, no BRCA1 or BRCA2 pathogenic variant (PV) was identified, only 2 MLH1 PVs.3 VYOC seems associated with MMRD-related EOC as opposed to homologous recombination deficiency (HRD)-related EOC as is seen in BRCA1/2 carriers, in whom the risk of EOC increases from 35 years for BRCA1 carriers and from 45 years in BRCA2 carriers.2 4

We retrospectively assessed the presence of PVs in VYOC cases aged <30 and 30–34 years before the main risk period is associated with PVs in BRCA1/2. These women had been referred to Manchester Centre for Genomic Medicine (MCGM) within the last two decades following diagnosis with VYOC in the North-West of England. The genetic testing described was performed as part of standard diagnostic testing within MCGM.5 A series of 77 women with ovarian cancer (9 borderline and 1 granulosa cell tumour in addition to EOCs) were screened for MMRD and HRD germline PVs by sequencing, multiple ligation-dependent probe amplification and a prescreen for MMR immunohistochemistry (IHC) as previously described.5 We also assessed the proportion of known carriers that developed EOC at age <30 and 30–34 years from our extensive dataset of >4000 female BRCA1/2 carriers and 910 MMRD heterozygotes.

Of the 77 ovarian cancer cases aged <30 years, no BRCA1/2 PV was identified. However, five MMRD PVs (four MSH2, one PMS2) were detected, making up 6.5% of all cases and 7.5% of epithelial cases, as shown in table 1. Of the 69 invasive tumours, 5 were of unspecified histological subtype (1xMSH2), 2 clear cell (1xMSH2), 7 endometrioid (1xMSH2, 1xPMS2), 29 mucinous and 12 serous tumours not otherwise specified (1xMSH2). Age range was 15.0–29.9; mean age=25.00, median age=25.77, IQR=23.23–28.00. The age of the four MSH2 heterozygotes was 23.6, 25.1, 26.2 and 27.5 years and that of PMS2 homozygote was 26.8 years.

Table 1

Breakdown of VYOC cases by histological subtype

When assessing the proportions of VYOC in PV carriers of HRD and MMRD genes, we included all tested and obligate carriers. There were 2005 female BRCA1, 1999 BRCA2 and 393 MSH2 PV heterozygotes (table 2). One BRCA1 PV carrier (obligate carrier) was identified with an EOC at age <30 (0.05%) (26 years old at diagnosis) but pathology subtype was unavailable. In contrast to the low rate in BRCA1/2 carriers, 4 out of 393 VYOC cases (1%) were found to carry the same MSH2 PV. This proportion was significantly greater in MSH2 PV carriers than either BRCA1 (p=0.003) or BRCA2 (p=0.0007) by χ2 testing.

Table 2

Proportion of women with VYOC with heterozygous PVs in BRCA1/2 and MMRD genes

Of the 2005 BRCA1 and 1999 BRCA2 carriers, six and three cases of ovarian cancer were diagnosed, respectively, between 30 and 34 years. This was significantly less than the five cases found in 393 MSH2 PV carriers (p=0.02; 0.004, respectively). The proportion of ovarian cancer in MSH2 PV carriers <35 years was significantly higher than 0 out of 278 found for MLH1 PV carriers. In addition to the previous four cases, there were two clear cell, one endometrioid, one serous and one yolk sac tumour (non-EOC, but with MSH2 loss detected by IHC in tumour) histological subtypes. Only one MSH2 PV carrier died from EOC with 77% surviving 10 years and 66% surviving >15 years.

In contrast, 7 out of 10 BRCA1/2 PV heterozygotes had died, with 4 died <5 years of diagnosis. Two long-term BRCA2 survivors both diagnosed at aged 32 had mucinous tumours (the remaining histologies were high-grade serous: BRCA1=2, adenocarcinoma not otherwise specified BRCA1=5; endometrioid BRCA2=1), raising doubts as to whether these were HRD-driven tumours. Hypothesising that these mucinous cases were not HRD-driven, survival at 12 years was significantly better in MSH2 (77%) than BRCA1/2 heterozygotes (15%; p=0.01).

Very few studies have addressed the contribution of HRD and MMRD genes to VYOC. In addition to the study of 101 cases aged <30,3 we identified a study of 47 women diagnosed at age ≤40 with EOC who underwent germline screening for 11 genes associated with ovarian cancer. This identified PVs in 13 (28%) of women (BRCA1: 10, BRCA2: 1, MSH2: 1, RAD51D: 1).6 This study included only two women diagnosed under 30 years of age, neither of whom had an identifiable PV.

Our study has shown that while the genetic predisposition for many early onset ovarian cancers is still unknown, MSH2 is the most important EOC predisposition gene at age <35 years. The cumulative likelihood of an EOC in MSH2 heterozygotes would appear to be >2% by 35, with this likelihood still below 0.5% for BRCA1 and rare for BRCA22; indeed, two-thirds of cases identified in BRCA2 carriers may not have been driven by HRD. This increased incidence despite the good long-term survival in MSH2 should prompt awareness of the increased risk and consideration for early risk-reduction strategies.

Ethics statements

Patient consent for publication



  • Twitter @drnikiflaum, @DrEmmaCrosbie, @ER_Woodward

  • Collaborators Not applicable.

  • Contributors The project was conceived by DGE and NF, and primary manuscript was written by NF. Statistics was performed by DGE and NF. All authors commented on and edited the manuscript.

  • Funding DGE and EJC are supported by the National Institute for Health Research (NIHR) Manchester Biomedical Research Centre (IS-BRC-1215-20007). EJC is an NIHR Advanced Fellow (NIHR300650). NF is supported by CRUK via the funding to Cancer Research UK Manchester Centre: (C147/A18083) and (C147/A25254).

  • Competing interests None declared.

  • Provenance and peer review Not commissioned; externally peer reviewed.