Article Text

Download PDFPDF

Next-generation sequencing-based BRCA testing on cytological specimens from ovarian cancer ascites reveals high concordance with tumour tissue analysis
  1. Caterina Fumagalli1,
  2. Alessandra Rappa1,
  3. Chiara Casadio1,
  4. Ilaria Betella2,
  5. Nicoletta Colombo3,4,
  6. Massimo Barberis1,
  7. Elena Guerini-Rocco5
  1. 1 Division of Pathology, IEO, European Institute of Oncology, IRCCS, Milano, Italy
  2. 2 Division of Gynecologic Surgery, IEO, European Institute of Oncology, IRCCS, Milano, Italy
  3. 3 Division of Gynecologic Oncology, IEO, European Institute of Oncology IRCCS, Milan, Italy
  4. 4 School of Medicine and Surgery, Università degli Studi di Milano-Bicocca, Milan, Italy
  5. 5 Department of Oncology and Hemato-Oncology, University of Milan, Milan, Italy
  1. Correspondence to Professor Massimo Barberis, Division of Pathology, European Institute of Oncology, IRCCS, Milano 20141, Italy; massimo.barberis{at}


Background With the approval of the poly (ADP-ribose) polymerase (PARP) inhibitor olaparib for newly diagnosed, breast cancer gene (BRCA)1/2 mutated, ovarian cancer women, the assessment of BRCA1/2 tumour status will be shortly required at the time of diagnosis.

Aim To investigate the feasibility of next-generation sequencing (NGS)-based BRCA tumour test on cytological specimens from ovarian cancer ascites.

Methods We evaluated the BRCA1/2 status on neoplastic ascites and corresponding tumour tissue of 11 patients with ovarian cancer, using the NGS ‘Oncomine BRCA Research Assay’.

Results The NGS-based BRCA test on cytological samples had a success rate of 100%, with 11 of 11 concordant BRCA1/2 results between ascites and tumour tissues analyses, including two wild type samples and nine cases harbouring somatic or germline variants.

Conclusion BRCA test may be performed on ovarian cancer ascites, reproducing BRCA1/2 tumour status and representing a useful tool for clinical decision-making.

  • ovarian tumour
  • cytopathology
  • molecular genetics

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.


Epithelial ovarian cancer (EOC) is an aggressive disease with poor clinical outcome,1 frequently diagnosed at an advanced stage with evidence of widespread peritoneal carcinomatosis and malignant ascites.2 poly (ADP-ribose) polymerase (PARP) inhibitor (PARPi) therapy has improved the clinical outcome of women affected by platinum-sensitive recurrent ovarian cancer, in particular for those harbouring BRCA1/2 mutations.3–5 Moreover, according to the recent results of SOLO1 trial, the Food and Drug Administration approved the PARPi olaparib for patients with newly diagnosed BRCA-mutated (germline or somatic) advanced EOC who are in complete or partial response to first-line platinum-based chemotherapy.6 7 Therefore, the assessment of the BRCA1/2 status at the time of diagnosis of ovarian cancer and the early identification of BRCA1/2 alterations are pivotal steps for clinical decision-making.

Peritoneal washing is routinely performed during ovarian cancer surgery for staging purposes. Moreover, malignant ascites can be drained before surgery to relieve patient pain and discomfort . These fluids may represent a source of ovarian cancer cells that could be characterised not only phenotypically but also evaluated for molecular alterations, including BRCA1/2 mutations.

BRCA testing is feasible on whole blood samples, evaluating germline BRCA1/2 status and on tumour tissue, investigating both germline and somatic BRCA1/2 variants.8 In this study, we (1) assessed the feasibility of BRCA1/2 analysis on cytological specimens from malignant ascites, using next-generation sequencing (NGS) and (2) retrospectively evaluated the concordance of BRCA1/2 status between cytological and matched tumour tissue samples.

Patients and methods

Patient population

This feasibility study included 11 patients with EOC, who underwent surgery and ascites drainage or washing prior to or during the surgical procedure at the European Institute of Oncology. Tumour BRCA1/2 status was available for all the patients, as tumour BRCA test was required by gynaecologic oncologists after the histological diagnosis of non-mucinous and non-borderline EOC, according to the test guidelines.9 Each patient gave written informed consent to perform tumour BRCA test. The clinicopathological characteristics of the patients were summarised in table 1.

Table 1

Clinicopathological characteristics of the patients included in the study

Ascitic fluid sample preparation and DNA extraction

Cytological specimens were retrospectively retrieved from the archives of the Division of Pathology of European Institute of Oncology. Ascitic fluid samples were prepared with cytocentrifuge (cytospin smears) and stained with May-Grunwald/Giemsa (figure 1). The identification of ovarian malignant cells in peritoneal fluids was performed by cytopathologist. The most representative cytological slide for each case was selected for the analysis. The slide was unmounted, and the stained smear was scraped in a tube containing 100% alcohol. The DNA was then extracted automatically with the Promega Maxwell instrument (Promega, Madison, Wisconsin, USA) and quantified with the Quantus fluorimeter (Promega).

Figure 1

Cytospin smears from malignant ascitic fluid stained with May-Grunwald/Giemsa. (A) Clusters of malignant tumour cells (tumour cell percentage more than 50%); (B) isolated tumour cells spherule in an inflammatory background (tumour cell percentage less than 20%).

BRCA1/2 NGS analysis

The BRCA1/2 status was evaluated using the NGS panel ‘Oncomine BRCA Research Assay’ (Thermo Fisher Scientific, Waltham, Massachusetts, USA), following the manufacturer’s instructions. Both library preparation and the subsequent chip loading were automatically performed on the Ion Chef System (Thermo Fisher Scientific). The sequencing was run on the Ion S5 System (Thermo Fisher Scientific), and data were analysed on the Ion Reporter Analysis software (v. 5.10) using Oncomine BRCA (5.10) bioinformatic pipeline (Thermo Fisher Scientific), covering single-nucleotide variants and insertions/deletions (indels). Each variant was then visually inspected using the Integrative Genomics Viewer software. The variants were classified according to five class system of the IARC (International Agency for Research on Cancer) clinical classification, as pathogenic, likely pathogenic, variant of uncertain significance (VUS), likely benign and benign variant. The classification of each variant was defined according to the ENIGMA consortium revision on the BRCA Exchange database ( If the variant was labelled as ‘not yet reviewed’ by ENIGMA consortium, other public databases as ClinVar ( or Leiden Open source Variation Database ( were consulted. BRCA test results on DNA from cytological specimens were compared with the results obtained from the BRCA test previously performed on DNA extracted from the matched formalin-fixed paraffin-embedded (FFPE) tumour tissue.


BRCA test performance on malignant ascites

The percentage of tumour cells in the sample and the concentration of DNA recovered from the most representative cytological smear of each case were reported in table 2. The median concentration of extracted DNA was 0.9 ng/µL (range 0.1–1.9 ng/µL), eluted to a final volume of 60 µL.

Table 2

Percentage of tumour cells in the sample, concentration of DNA extracted from a single smear and NGS metrics of each cytological sample

The NGS run metrics obtained for each cytological specimen were reported in table 2. All the 11 cytological cases reached the quality parameter in terms of on target (>85%), mean depth (>1200) and uniformity (>85%) of mapped reads, despite the starting quantity of DNA of four samples (3, 4, 9, 11) was below the amount suggested from the NGS panel (10 ng of DNA in 15 µL for the library preparations). The turnaround time from DNA extraction to the BRCA test result was 5 days.

BRCA1/2 status on cytological and matched tumour tissue samples

The results of BRCA test on cytological specimens were reported in table 3. The concordance of BRCA1/2 status between malignant ascites and tumour tissue was 100% (11 of 11 concordant results), including two wild-type samples and cases harbouring BRCA1/2 pathogenic/likely pathogenic variants (n=7) or variants of uncertain significance (n=2). A slight difference in the variant allele frequencies (VAFs) was observed in cytological specimens as compared with tumour tissue samples. BRCA1/2 germline status was available for four patients. In two patients with no BRCA1/2 germline mutations, somatic BRCA1 variants were pinpointed both in the tumour tissue and cytological samples.

Table 3

BRCA1/2 status in FFPE tumour tissue and cytological specimens from ascites. When available, germline BRCA1/2 status were indicated


The BRCA1/2 status is crucial for the clinical management of women affected by EOC. Moreover, the presence of BRCA1/2 alteration, whether somatic or germline, may indicate a clinical benefit of PARPi treatment even for women with a newly diagnosed tumour. Therefore, an early evaluation of BRCA1/2 status represents an important clinical need.

The presence of malignant cells in ascitic fluid is a well-known factor of poor prognosis for patients with ovarian cancer that correlates with worse progression-free and overall survival.11 In the last years, different efforts have been made to exploit peritoneal fluids for the molecular characterisation of primary tumours.12–14 In particular, cancer cells,12 cell-free DNA13 and RNA extracellular vesicle RNA biomarkers14 from ascites have been thoroughly investigated. However, so far, few studies analysed BRCA1/2 status on malignant cells recovered from peritoneal/pleural fluids.15 16 Recently, Barquín et al reported on the feasibility of BRCA1/2 mutation analysis on 10 cytological samples from fresh peritoneal washings17 and Gornjec et al evaluated a large cohort of patients with ovarian cancer using capture-based NGS approaches.18 However, they assumed that fresh peritoneal liquid was available immediately for DNA extraction17 or used as a protocol for cytological specimen preparation including a cell medium addition step that is not routinely performed in most diagnostic laboratories.18 In the last years, our group demonstrated that archival smear slides may represent suitable samples for molecular tests, including NGS analysis.19 20 In the present study, we showed the feasibility of BRCA1/2 evaluation on archival stained smears from ascitic liquid drainage or peritoneal washing. The evaluation of each stained cytological specimens by an expert cytopathologist is a crucial preliminary step to assess the ovarian tumour cell content and reduce the risk of cross-reactivity with cells of different origin. Moreover, it may reduce the detection of clonal haematopoiesis-related genetic events, although no data on BRCA1/2 mutations and this phenomenon have been reported yet. We used an amplicon-based NGS panel and we achieved a BRCA test success rate of 100% despite four cases had a starting DNA yield below the suggested quantity. Indeed, as previously demonstrated higher quality DNA can be extracted from stained smear specimens.20 The BRCA1/2 status between FFPE tumour tissue and ascitic fluid specimens was concordant in all cases, including BRCA1/2 wild-type samples and cases harbouring BRCA1/2 somatic or germline alterations. Although we observed differences in the VAFs of BRCA1/2 mutations, these variances may be related to the different tumour cell content of cytological and tumour tissue samples.

Given the limited number of the cases included in this study, these results should be considered preliminary and to be confirmed in a larger cohort. However, these data may lead to the implementation of BRCA1/2 testing on cytological samples in a routine diagnostic setting. Indeed, the evaluation of BRCA1/2 status on ascitic fluid specimens may be clinically useful as: (1) peritoneal fluid drainage is a minimally invasive procedure, often prior to the surgery to relieve symptoms associated to the fluid build-up; (2) the BRCA1/2 status may impact on the surgical procedure, since the BRCA1/2 positive status was recently associated with a survival benefit in patients with gross residual disease21 ; (3) peritoneal liquid offers a dynamic snapshot of tumour genomic alterations, enabling to capture intratumour genetic heterogeneity as compared with single-site sample analysis and (4) both somatic and germline variants could be identified in malignant ascite cells as in tumour tissue.

In conclusion, we reported on the BRCA1/2 assessment on stained cytological smear from ascites using amplicon-based NGS. Although the number of cases is limited, the high success rate, the turnaround time compatible with clinical needs and the high concordance with tumour tissue BRCA1/2 analysis are promising achievements that suggest the feasibility and reliability of the evaluation of BRCA1/2 status on cytological preparation from ovarian cancer ascites.


The authors acknowledge laboratory technicians of Division of Pathology, European Institute of Oncology, Milan, Italy.



  • Handling editor Runjan Chetty.

  • Contributors All the authors significantly contributed to the ideation, laboratory analysis, data elaboration and final manuscript writing and editing.

  • Funding The authors have not declared a specific grant for this research from any funding agency in the public, commercial or not-for-profit sectors.

  • Competing interests CF received honoraria from Roche; EG-R received honoraria/advisory fee from Thermo Fisher Scientific, Roche, Novartis and AstraZeneca; and MB received honoraria from Thermo Fisher Scientific, Roche, BMS, MSD, and Biocartis.

  • Patient consent for publication Not required.

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